Sarah (00:08)
Welcome to UnDUBBED the podcast that's unscripted, uncensored and undeniably data. I'm Sarah.
Fiona (00:14)
And I'm Fiona, quick question. When did you last look at a dashboard and actually change something because of it?
Sarah (00:22)
If you're pausing right now, that's the whole episode.
Fiona (00:28)
Today we're talking about dashboards. Not how to build them, but whether we should.
Sarah (00:35)
we've got two guests who literally co-wrote the book on this. Actually, between them, they've co-written two books on dashboards
And they co-host a show together. So this is going to be either brilliant or a very organised argument.
Fiona (00:51)
Amanda Makulec is a data visualisation leader who spent 15 years helping teams communicate data effectively with a particular focus on health data. She's a founding board member and former executive director of the Data Visualisation Society and author of Dashboards That Deliver.
Sarah (01:09)
And Andy Cotgreave has spent 20 years in data analytics, 15 of them at Tableau,
Andy Cotgreave has spent 20 years in data and analytics, 15 of them at Tableau, and is now focused on how we communicate data and what AI is about to do to all of it. He's also the co-author of Dashboards That Deliver and the Big Book of Dashboards.
Fiona (01:33)
And together they co-host Chart Chat, a series on data communication. so they are genuinely used to talking to each other. We'll see if it makes our job easier or harder.
Sarah (01:44)
Amanda and Andy, welcome to UnDUBBED
Andy (01:49)
Good afternoon, evening. Evening for me. Yeah, great to be here.
Amanda (01:51)
So happy to be here.
Fiona (01:55)
it really is wonderful for you to make the time. I'm being a little bit cheeky. I know how bad the time zones are. Before we go any further though, this one's for the listeners.
Andy (02:02)
It's fine.
Fiona (02:08)
please prime those algos, hit follow, leave us a review and share this with your data team, especially the person who builds dashboards that nobody reads.
Sarah (02:16)
They're all out there.
Fiona (02:17)
you
Andy (02:18)
Yeah, yeah, yeah.
Sarah (02:20)
So we like to start with the person behind the profession. Amanda, can you tell our listeners a little bit about your journey, where you've come from and where you are now?
Amanda (02:30)
Sure, it's a delight to be here. I'm Amanda Makulec. I currently am an independent consultant doing data visualisation design, primarily still in the public health sector. But my journey started out with my master's in public health, going ahead and serving as an analyst and working my way through various different roles where collecting and analyzing data was a big part of my job. And data visualisation was a little piece of that.
As part of that journey, I spent a lot of time in Sub-Saharan Africa training and teaching teams on how to build effective charts and graphs and dashboards, as well as getting to spend a lot of time thinking about data demand and use. How do we actually get people to effectively want to engage with data and information and make sure they have the right data at the right place in the right time? All of that time I spent abroad is one of the big reasons I helped to start up the Data Visualisation Society, which helps connect data as practitioners around the world.
And being involved in that organisation still as an advisor is a huge passion project of mine because I feel like data is that kind collective threat that pulls people together across industries and across different spaces. So excited to dive in and debate with my chart chat chum and the rest of you all about the world of dashboard design and how we get people to actually use them.
Fiona (03:41)
One thing that popped into mind with all of your great experience is I'd really love for you to join our consultants and freelancers Tableau user group. We've got a whatsapp group. There's about 80 people on it now it's kind of lively and it focuses on people who
have their own business or, you know, are solopreneurs and, you know, really trying to make it in the way of data visualisation. So I'll make sure I pop that in the show notes and invite you along afterwards as well. Awesome. Thank you. And Andy, I know you've been here before, but in case people don't know who you are, say a question for you. What's your journey?
Amanda (04:08)
Would love that.
Andy (04:20)
My journey, all right, let's go super fast. Between school and joining Tableau, I was trained or aspired to be an artist, a geographer, a glaciologist, a software engineer, a usability researcher, a business researcher, a cycle guide, and then became a data analyst, right? And then I became a data analyst at Oxford, discovered Tableau, and joined Tableau in 2011.
Right. The reason I say all those things is because those skills in my early part of career were design, programming, coding, empathy with human beings, usability, and all the skills one needs to be to be a data journalist. Then I joined Tableau, became the evangelist there from 2016 onwards, joined in 2011 and had just the best job in the world, spreading the mission to help people see and understand data. But then since I last was on the show, I left Tableau in
September 2026 to co-found my own company, How to Speak Data. And now I have the mission to teach people how to speak data through training, keynotes, more books, and just looking at the application landscape and seeing what the heck is happening in this period of crazy disruption. So yeah, great to be back.
Sarah (05:32)
Love your journey, Andy, and I'm sure lots of people would really resonate with all the elements that you kind of bring together and that well-roundedness of kind of starting in the trenches and working all the way up to the evangelism and now running your own business. So cool. Yeah. Now you two already know each other pretty well, being co-authors and co-hosts. Is there anything about each other's work
Andy (05:48)
Yeah, it's fun.
Sarah (05:58)
that still surprises you. I'm gonna pass this one over to you, Amanda.
Amanda (06:02)
Well, I mean, it surprised me that Andy's first deep dive into the world of solopreneuring included pumping up companies that were not Tableau. That was a change. And I think that one of the things that does not surprise me about Andy is I think that no matter where he goes, he is someone who is a champion for effective data communication for people of all technology skills and levels. And so I've appreciated the work he's done tinkering with different kind of AI enablement tools.
Andy (06:14)
You
Amanda (06:30)
surprises me though is that I feel like Annie for a long time you were sitting over on the side of the like not so fond of the AI fence. And I feel like you've slowly started climbing over the fence over to where Mr. Jeff Shaffer sits while Jeff Shaffer climbs back my direction. So we will see what that looks like in the long arc of time.
Andy (06:47)
Yes,
indeed. That's very flattering. Thank you, Amanda. I'm touched. Yeah, yeah.
Fiona (06:53)
that's really nice, isn't it?
Amanda (06:55)
I mean, you said you were still
going to write more books in your intro, Andy. That means you haven't been scarred from the past two with the three of us. I'm flattered by that.
Andy (07:04)
Can I reciprocate and answer the same question about Amanda? All right. So we we spent Amanda and I particularly spent a lot of time during the writing of dashboards to deliver working together. I think Amanda's
Sarah (07:06)
course.
Fiona (07:06)
Yes.
Andy (07:16)
knowledge of the project, whole before you get to development phase of building dashboards. I learned so much from Amanda about discovery, prototyping, and the way she developed incredible, well, she developed the framework that we feature in, dashboards that deliver. And I just wish I had that when I was a data analyst at the University of Oxford.
So I think Amanda's passion for that skill for that is fantastic. Amanda is a brilliant networker as well. having been an outlier in Miami with her, the DataVis Society event last year, just seeing her work the people, she's just brilliant because she's charismatic, charismatic, passionate and enthusiastic. So β it's a pleasure. Well, I...
Amanda (08:02)
I just find other people so interesting, Andy. It's authentic. I really do.
Fiona (08:06)
Yeah.
Andy (08:07)
I mean, in some ways I'm similar, right? I think in that aspect, we're cut from the same cloth, right? People are fascinating. And so it's just great getting out and listening to people and seeing them and to do the same. The AI thing, holy cow. Yeah, we will have to come to that because, yeah, we'll save the AI conversation for a little bit later, but it's fascinating.
Amanda (08:18)
Thank you.
Sarah (08:27)
Well, thanks for answering both of those questions. I think it's a lot to unpack there and we'll go through some of that later on.
Now, given all the work that we all do in data, why not delve into the uncomfortable question? Are dashboards as they exist today in most organisations actually delivering value?
Fiona (08:51)
And a
Andy (08:52)
Well, we did a survey during the production of our book. We were like, let's ask our network a simple question. Do most dashboards fail? was, so we asked that question of our, our combined network, our re all of all our followers across channels and 71 % of people answered that question. Yes. Most dashboards fail. We didn't define the criteria on which they should judge the answer, but that is a large percentage.
And somewhat depressing since a significant number of our network probably bought the big book of dashboards nine years ago and that didn't solve the problem. So yes, a huge amount of dashboards fail. think bringing Amanda on board for dashboards to deliver and you know, thinking deeply about why that is, is because we aren't thinking deeply enough about how we serve our users. And so again, as I alluded to a minute ago, the framework that Amanda brought to the book,
is really what's key to trying to address that issue. How can you get the users on board early in the process so that they adopt and engage the things you build? So yes, we agree. Lots of dashboards fail. And the reasons are many and complex. Hopefully, we've addressed some of them in dashboards that deliver.
Sarah (10:09)
It's an interesting point that you bring up. I still get surprised at how many dashboards are built in silo without the users and constantly trying to get our clients to think of this differently.
Andy (10:17)
Mm.
so when I was at the University of Oxford, I'd read my Stephen Few books. I was a pro at Tableau and I went, what do you want? You want to know about student progression through university? Go away, I'll come back in two months. Ta-da! And they hated it. I was like, why do you hate this? Look, it's got no pie charts. It's got beautiful bar and bar charts and every has got 25 filters on it. Come on, losers. And then it's only with many years hindsight that I realized I was the loser because I did.
really pay attention
Amanda (10:50)
I'm going to, now I agree, I think Andy's stats are right and I will caveat because I had my roots in research, like I said, there was a survey that included 465-ish responses for a bit of a sample size context there and people who are in our network who are dashboard developers. And yes, many of them said that most dashboards fail. So the 71%. I think the two things that we bring up though in the book and the two things I index a lot in my mind when I think about why do we still build these kinds of data tools is,
Andy (11:01)
Mm-hmm.
Ahem.
Amanda (11:18)
Do you have a clear sense of what success looks like from your dashboard in the first place? And how are people defining fail when we say, do they fail? That's a big, bold statement. And so do you have a clear vision of what success looks like? How are you going to measure it? And I think almost more importantly, I think a lot of the dashboards we think of at an enterprise level are the big deployed dashboards, right? They're dashboards that are out there on servers and in other spaces.
And I think if you dig in and start to think more broadly about dashboards, by the definition of what we have as a visual display of information meant to monitor conditions or facilitate understanding, I think there are a lot of data tools people build for themselves and have on their own spreadsheets and other spaces that function as dashboards because people still need data and information. They're just not dashboards that get captured in that kind of big headcount of users and page views and all of those other pieces.
And the value proposition for dashboards, the changing minds, and you said at the beginning of this, right? How often do you make a decision or change something because of a dashboard? It's hard to quantify and count those things because someone has to follow up on a survey or some other way of tracking when those things happen. And so I think quantifying success is actually one of the harder parts of making the use case value proposition for what dashboards are and can do. And so I'm going to say Andy is right on the step.
But I think there's some context that's worth wrapping around that.
Fiona (12:43)
So digging deeper on that, I totally agree quantifying it is super important. And you mentioned, well, you could do a survey or there's other ways. Can you give us some examples where you've seen people quantifying things to measure the success and then anything that they've done, say they fail, if they're 71 % of failing, like what are they doing then to adjust it? How do they approach that?
Amanda (13:12)
man, my research hat is back on. 71 % of people agreed that most dashboards fail. That's people as the end, not 71 % of dashboards fail. So I want to be fair to that and what that looks like. I think when it comes to measurement, We talk in our book about outputs and kind of having measures of success that are small s success, the countable stuff.
the page views and counts and really using those directionally to look and see where do people stop looking at a dashboard and do we need to revisit why someone's not looking at the dashboard anymore and it's not getting any page views? Should it be archived in some way, or form or updated to be more usable for someone? And then there's the big success metrics that are really the business impact side of things that showcase kind of how something gets used to make a decision. And I found in my career, especially in public health,
Those big S things, the impactful moments where someone sees a data point on a dashboard and then uses it to do something, happen not when someone's sitting at their desk looking at a dashboard, but when a dashboard is being brought up in real time in a meeting or another space, enabling conversation and a starting point to talk about so that the people in the room can talk about why that data point has changed, has gone up or down, why this trend looks a certain way. And those conversations are where the
the value and impact happen, which is why think dashboards still have a role to play even in a very AI obsessed world.
Fiona (14:38)
Yeah, we're hearing about it a lot at the moment about having analytics and data in the workflow. So at the moment, it's very external to the workflow. You have to log on to the server or the cloud or wherever you're logging into, go and take a look at it. And it's not integrated where people are actually doing their work. Have you started to see
a lot more of that integration or do you think that it's all just people saying that would be the nice thing to happen?
Andy (15:07)
So.
Let me tell you two very quick anecdotes related to that. When I was at Tableau, we were working with one of the big government departments in the UK. They showed a chart of how they'd been building pilot dashboards to get their civil servants used to Tableau and data. They were seeing more and people logging onto the Tableau server. They were beginning to build the second generation of more action.
actionable dashboards. And then views fell off a cliff. And it was because the IT department saw this new server and went, well, we'd better put two-factor authentication on that. And I don't have a problem with governance, but it killed the project because two-factor authentication was too much friction. And there's no easy solution to that because governance is important. But if you govern the actual existence of a project,
away, then something's not quite right with that security protocol. So that was a real problem about friction. And then there was a chapter I wrote for the big book of dashboards which never got into the book. And that was a dashboard by a company that does emergency counseling in the US by text message. And the dashboard was just a text. Text characters served as a message in Slack.
So there were no graphics, it was just numbers, it was just the KPIs and a bit of text, and anybody could call that dashboard on demand. And I thought it was brilliant, because the guy, Bob Philbin, who's the CDO at Crisis Text Line, recognized that if you're just writing some scripts to serve text, then you don't have to pay a license to a visualisation firm, so there's lot of maintenance, and it's a super simple means of delivery that is in the flow of people's work. So I think it was a genius.
move to reduce friction and put the data where people were. It was a bit too radical, I think, for Steve and Jeff to go in the book. So we pulled the chapter, but it remains a really great epitome of getting data where people want in the mode with which they can respond to it very, very quickly without leaving their flow.
Fiona (17:10)
I think that's something that you could really run with with your business as well, like helping people to get data into their workflow.
Andy (17:13)
Yeah.
Yeah, absolutely. people are like, oh, there's no graphics in it. Do you remember we were all obsessed by Wordle about three years ago? One of the reasons Wordle succeeded was because you could copy and paste text into a WhatsApp message. Those little green and yellow things were just Unicode characters. So we were pasting a text-based dashboard to billions and billions of people every single day because it was low friction. It's brilliant. So there you go.
Sarah (17:19)
Mm-mm.
Mm-hmm.
Andy (17:48)
Everybody ditch all your BI tools. All dashboards should be Unicode. That's what I've learned.
Amanda (17:53)
Andy, is like, this is like, have a whole, we have a whole section on how
we have to have context with our bands. And it sounds like you're saying we're going to automate the bands without the context, but the friction point is a very good one.
Sarah (17:57)
Hahaha
Andy (18:02)
Well,
there are Unicode arrows, right? You can put it up and down. can add context with text.
Amanda (18:06)
There are, directionally, yep.
I mean, to Andy's credit, he really loved this dashboard so much that he brought this up when we were writing the new book about debating if this was a dashboard now and if he could convince everyone to be on board with it being a dashboard. But comically, there was the whole debate about should we have a chapter on a NASA hyper wall that is a 20-foot long display in an Earth information center at NASA headquarters, now replicated at a few other locations, including museums.
Andy (18:26)
Yes.
Amanda (18:35)
that displays information for people on demand, kind of in a scrolling loop setting, in a space that people could wander into and view and explore. And it wasn't meant for decision making, but rather awareness, engagement, conversation. And we had a whole debate about if this kind of dashboard was even relevant to talk about for the audience of our book in terms of analysts and developers. And I stand by that it's one of my favorite chapters and scenarios in the book because it
puts data where people are and is such an interesting public facing use case where so often people say they're designing dashboards for the general public and it's meaningless and useless, but really NASA was designing dashboards for the general public and policymakers in this very specific setting.
Andy (19:22)
I'll add one other thing that Amanda doesn't mention there. It's also one of my favorite chapters in the book is that both Big Booker Dashboards and Dashboards that deliver are packed full of real world scenarios, right? And so the thing that's great about the NASA dashboard as well is that even though you might not be building a 24 massive wall for thousands of people, members of the public, that might not be your use case, but what NASA did in terms of their process.
Sarah (19:22)
Yeah.
Andy (19:49)
was everything you would do if you were building a KPI dashboard for your CEO and their board, right? So it's just a really good chapter, A, for everything Amanda said, but also because of the process. It's something we can all.
Fiona (20:01)
Mm.
Sarah (20:02)
Nice. And I think, you know, when I think about the kind of bite size dashboards or being in the flow, I think of like the whole 'Tik-Tok-ification' of everything as well, right? We're used to consuming things in such small pointed bites now. And that kind of leads me to think is that where we're as the, is it the gen, where are we up to now? Gen alpha or something? Yeah, start.
Fiona (20:29)
Gen Z to Alpha, yeah, coming in.
Andy (20:29)
think so.
Amanda (20:29)
Plus,
Sarah (20:32)
start coming in. So Andy, you've seen thousands of dashboards across your career. What's the most common mistake people make before they even open the tool?
Andy (20:42)
The most common mistake, my answer is it goes back to that, the framework we wrote. They dive into development too early without thinking about what users needs. And so the first two pieces of our framework are about discovery, then prototyping. And we borrowed the double diamond from the design council, which is, which I'm not gonna try and mine, but essentially,
You expand to get as many ideas in discovery phase before narrowing down to the scope of what you can cope in the project. And then you widen out again as you prototype loads of ideas, throw loads of things that users get, or get the users to throw loads of ideas around. So widen out and then narrow down again as you get into development. And I think that just crystallizing that was hugely important for me as I, well, I don't develop so many dashboards these days.
But I think, were I to be doing a dashboard project today, that would be the biggest thing I learned.
Fiona (21:37)
why don't you two build a course on it? So you've got the book, but there's one thing in reading to learn about something, but then there's another thing about actually putting it into practice and having something that's a bit more tangible. For me, like that whole muscle memory is super important. Have you considered doing something like that? Because I think it is a real gap in market.
Andy (21:54)
Yeah?
Amanda (22:02)
I actually teach a two day workshop on it. I ran one in Washington DC in the fall in terms of an in-person workshop and I'm currently adapting it into a cohort based course on Maven. So that'll be up live sometime this spring-ish. So we are doing that or I am at least. And I think each of us teaches it in different ways and spaces. I know Jeff uses the book in his own instruction and we've seen other professors who have picked it up, which has been, I think really a joy to see it being influencing people earlier in their career. I know I don't have Andy's tenure or gray hair.
Fiona (22:12)
nice.
Woohoo!
Andy (22:19)
Mm-hmm.
Amanda (22:31)
But I think that as we go ahead and kind of look at what this book could have meant for me when I was younger, it's the book I wish I had when I started my data viz journey. I think it's interesting, Andy, you somehow managed to answer a question about dashboard failures and big picture frameworks and never said the word user or audience at all. And in my head, I always go back to the fact that when we go ahead and we design dashboards, if we design dashboards around job titles as our users instead of people, we often miss the boat.
Because we focus entirely on just what their functional requirements are in their job, rather than how they actually like to engage with data and information, what their graphicacy is, how they're used to seeing the data and information, where they go for it, what decisions they can make with it. And it's interesting because I think, Andy, you're always the first one to jump into talking about audiences and users, feel like, but you managed to give a whole answer without it.
Andy (23:23)
Well, I was just leaving the door open for you, Amanda.
Amanda (23:26)
I appreciate that. They don't know what's in
our framework, right? I think it's an interesting point on the market gap though, in terms of dashboard instruction. Because I do think we thought about that with the book too, right? That there are so many books that have kind of pieces of agile software development or project management or looking at chart and dashboard best practices through a very kind of...
Fiona (23:29)
Hahaha
Amanda (23:48)
very specific flow charty kind of lens, like what Nick DeBarra does, that are all really valuable compliments to this kind of piece. But that framework piece of how do actual teams working within real world constraints build things is really important.
Fiona (24:02)
I would like to sort of dive into now the future of dashboards.
take a deep breath, something that we've all been working on for a long, time. Are we ready to leave it behind? Do we need to? Why don't we dig into the teaching directly? There's a growing conversation about whether dashboards are becoming obsolete. Thought spot, dashboards are dead. And whether they will be replaced by AI generated answers, conversational interfaces or automated insights.
Amanda, let's start with you. Where do you land on that?
Amanda (24:40)
I think that we'll continue to see evolution in how people engage with and interact with and source data. Gen.ai tools are one piece of that, but we've seen that evolution happen over a very long arc of time, all the way through Excel and into the world of different old school BI tools, the kind Cognos business objects space and into more visual displays of information. And I think there's something good about having more tools out there that make it easier and reduce the friction.
for people to go ahead and access the data and information they need. Whether it's on something called a dashboard, a live board, I think we saw algorithmic cockpit in our research about what people were saying dashboards were these days. But the need still is there, right? People need information and I think that AI tools will become one bigger part of that toolbox. We've seen that in the DataViz Society's annual State of the Industry survey, that people doing data viz development continue to increase the amount
Andy (25:18)
You
Amanda (25:36)
or the share of people who actually use AI tools in their work. I don't think dashboards are dead though, because we still need access to information. How we create them might look a little bit different. And maybe we can automate away, away is the wrong word maybe, maybe we can automate some of the simpler parts of the process that allow us to go ahead and get to those working prototypes faster. But I think there's value in the prototyping part that I worry about us losing.
A big part of prototyping is actually showing things to other people and getting their feedback, not just collaborating with the chat bot on what's kind of the best practice in your data visualisation. And so I think there's still a big need for human designers and developers to be creating the space for those conversations and getting feedback from real users. And I think that we still are at a place where there is a lot, there are a lot of challenges that are related to people, process, governance, definitions of different words.
that if I say conversion rate to someone within a large organisation, that might mean really something very different for someone who works in the digital space versus someone who works in a business operation space versus someone who works in finance. And until we can automate away some of that resolving those differences of definitions, you could be serving up very nice looking data that functionally is wrong. And I worry that the pretty data that's actually wrong
Andy (26:57)
Mm.
Amanda (26:59)
is the thing that's going to do a lot of us in, because you can break trust in data really quickly after you've spent years to build it up.
Sarah (27:06)
Yeah, it's so interesting, isn't it? Like you can, like what you've just said, it takes ages to build up trust and it can just be knocked down in one minute. How many boardroom meetings have we been in? And it's like, I don't trust that number. And you're presenting like 50 numbers and credibility gone.
Amanda (27:21)
Or worse,
you get into the boardroom and your CMO, your CTO, and your CEO are all bringing up the same metric with the same name from the same period. And because when you go back to the weeds of it, the logic in SQL or wherever else that you've written the code is different, the numbers are different, and suddenly you have three numbers that should match and they don't.
Andy (27:39)
I think one other aspect, Amanda, didn't really touch on though, is that the industry hype is moving away from dashboards. And I think that is probably impacting what
people are thinking they certainly leadership thinks they should be doing. So there's two aspects of that that are important is part of me thinks we could have named our book something more to do with data applications. It's like, okay, let's stop calling them dashboards. mean, dashboards is just semantic drift left over from stagecoaches, right? It's an arbitrary word. So we could call them data applications, right? And that'd be a bit more modern. And I do think though,
AI offers new opportunities to interact with data where you don't need that instant answer. And so that is going to be a transition just in our workflow and what we produce. That's slightly what Amanda said. And I was at a training course today with one of the newer kids on the block. And I was chatting to some customers there, a big tourist organisation in the UK. And they were rolling out this AI first data tools. And
I was like, what's your policy about whether those answers they create are right or wrong? If somebody takes the output of this tool to the board and they make a decision it's wrong, who's responsible? They're like, we haven't thought of that. We just roll in with the punches. well, that feels like what everyone's doing. But some company is going to make some major decision based on some data that came out of an AI driven tool and it's going to be wrong.
And who takes the flak? I bet it won't be the chief data officer. It'll be the data analyst first, but it should be the CDO. There you go.
Amanda (29:16)
And there's health care.
But I
mean, healthcare is an interesting space to watch for that because there are conversations happening now about if you're using AI enablement tools or decision support systems, how are we regulating those in the same way that we go ahead and we regulate medical devices? Because they functionally have the capability of creating bigger problems, challenges and poor outcomes than some of the medical device components that we might use that are external to our body. And so there's interesting discussions there around what that looks like.
Andy (29:25)
yeah.
Mmm.
Amanda (29:50)
And I think around this overall risk framework, we have to think about the risk of getting it wrong. So in a space where I'm going ahead and I'm building a kind of summary KPI and I'm off by a few thousand dollars of what our revenue was for last month, is there any kind of really big negative thing that's going to happen? Probably not if my numbers are directionally correct and overall look pretty good.
Andy (29:56)
β
Amanda (30:11)
something Jeff and I have talked about this a bit and something he's talked about is the importance of like people being willing to put their signature on something and saying this is human verified by me. And I'm putting my name against it to say that I have verified that these numbers are as correct as they need to be or can be. And I am putting my stamp of
the Andy Cotgreave or Amanda Makulec stamp of approval on things. And that might help with kind of the conversations about accountability. Charlie Burns, who's a very funny Wisconsin guy, jokes that a bratwurst is more regulated in Wisconsin than AI data centers are, which were a big hub for the AI data centers here in the U.S.
Andy (30:37)
Even that's really hard,
So if I'm a data analyst, I about the data analyst and the chief data officer relationship, right? If I'm a data analyst, then I guess it's me whose name should be on there, right? But then am I being given a workload that is manageable? Because if I put my name on the fact that this vaccine is 96 % effective and it actually is only 42 % effective, I mean, that's probably quite an extreme example.
and I've put my name on it, but I only had one hour to validate six hours of AI driven work, then actually it's not my fault. It's my manager's fault. It's the business organisation's fault for making me work too hard. So then the CDO's name needs to be on it. well, I'm just a minion to them, right? Do they have to validate it? So yeah, somebody's name needs to be put on it, but that's still not a panacea, is it? They're still fraught with risk. Yeah.
Amanda (31:35)
I think that's where the risk framework is, right? So like if the
risk of getting it wrong is really high,
There's a review and approval piece and when the risk of getting it wrong is high, that should be the case to be clear even for non-AI big data things, right? It's the reason we have peer review processes when we're publishing in journals. It's the reason we have
Andy (31:40)
Mm.
Amanda (31:54)
other processes in health and healthcare for doing testing and reviewing data. But I think that in AI, the ability to kind of hand off the blame and pass the baton and the likelihood that things could go awry without you realizing it is higher in my mind.
Andy (32:11)
Mm.
Sarah (32:11)
Mm.
Fiona (32:11)
Yeah, for sure. mean, recently I saw something on Reddit. The person posted that for the last three months, their whole exec team had been reporting off AI data and it blew up like this whole thing blew up and the original thread got deleted.
Andy (32:23)
Yeah.
Amanda (32:24)
Thank
Fiona (32:27)
the author deleted it, but people had snapshotted it and then it went through LinkedIn and a whole bunch of things. Whether or not it's really true, you can see how easy it would be for that to happen.
Andy (32:40)
Yeah.
Sarah (32:41)
It also surprises me how little we hear about the failure stories because I would say there's a lot more out there that we don't know about because people are taking massive risks with data.
Andy (32:50)
Wait.
Yeah,
and it's a lot easier to write something hypey and grifty on LinkedIn and make it seem like the world is changing. mean, God, I hope my content on LinkedIn doesn't come across like that. That's always a big risk. But it's easy to write, AI did this incredible thing, rather than AI cost me my job because I put the wrong numbers out. So I think we get
massively influenced by the hype machine.
Sarah (33:20)
I'm just going to switch gears a little bit because we just spent a lot of time in the AI space and we could spend probably all day talking about that. Both of you have spoken and written extensively about data communication, not just data visualisation. What's the difference and why does it matter?
Andy (33:38)
Steve talks a lot about the difference between story finding and storytelling. And I think that's a really good model. Dashboards and the analytical. So the exploratory process and dashboards as an asset you might produce are about story finding. It's like, is the data telling me? I'm gathering insights. And then you tell the stories, whether you tell them as the analyst or you provide stories that other people can
tell, or you provide stories that they can tell. They are completely different paradigms of design and approach. Story finding is about detail and nitty gritty small numbers, like literally physically small things. And you can get lots of detail on a screen. But the story tell is about what's the insight and how do I get that insight into the audience's brain in the amount of time they have to look at that slide or.
story or article in a newspaper or whatever the story is. So I think that that's the opening answer to that. It's about story finding or storytelling a different paradigms. Amanda?
Amanda (34:42)
Steve picked up
and has talked about that a lot more. He took that from an article I wrote for Nightingale in like 2022 or 2023. But Steve has run with it and he's taught it way more and he wrote the chapter on it on the book. So he gets all credit there for amplifying that call a little bit wider. But there's a Nightingale article about that. And that if someone wants to dive into that on the distinction between like story finding and storytelling without, you know, diving into the great chapter in our book, you can go and find that article on DBS's Nightingale.
Andy (34:48)
Amanda's idea of story finding and storytelling, I... Yeah.
Sarah (34:50)
Hahaha
Andy (34:57)
He shouldn't get all credit. He shouldn't get all I'm sorry, Amanda. Right. Yeah.
Amanda (35:10)
I think that the data visualisation, data communication piece is deeper than just kind of, we presenting data to explore or to explain an idea? Because to me, data visualisation is one tool in the broader toolbox of how we help people understand information. And oftentimes good data communication doesn't have to or even involve data visualisation. It's giving people one statistic, one piece of information at the time and point that they need it.
And it recognises that there are ways people engage with data that aren't aligned with the kind of rules of precise, quick to insight visualisations that we often celebrate. And so I think data communication helps us actually celebrate all of the things that we might know make a good chart, a good data visualisation effective, but the things like the graphic legends and all the text annotations and the good headline and all those bits.
that are about communicating information effectively. And so this kind of comes back to my background and where I spent a lot of my time, which was focusing on how do we get people to actually use data and information more. And almost never in my global health career being funded under a project by a work stream that was called data visualisation. It was always data demand and use or knowledge acquisition or knowledge sharing or knowledge management or something else that was around how people engage with information.
I think it's a really exciting time to think about those things. And I want to give all credit to Jason Forrest, who's starting up a whole new master's program at the School for Visual Arts in New York City. And he's defined it and called it the Data Visualisation and Communication Master's Program because of the importance he's been placing on that piece of communicating and engaging with data. And he has a great Fast Company article actually on that same topic.
Andy (36:36)
Mm.
Fiona (36:58)
Maybe you can share that and we can put it
You both have of touched on this at different points throughout the conversation. And it seems like we're staring people away from the word dashboard. And perhaps if it disappears, what could we replace it with? And Andy, you sort of said, is it an app? What are some of the things that come to mind from both of you?
Amanda (37:20)
Algorithmic cockpit for sure, algorithmic cockpit.
Andy (37:21)
Yeah, so
that comes from, yeah, that's Cathy O'Neil who wrote a great book called Weapons of Math Destruction. She's an advocate for better AI regulation and her consultancy will audit your firm and create algorithmic cockpits. There's a chapter in the book called What is a Dashboard? be honest, I don't give a shit what you call it.
Fiona (37:24)
That's so phallic and it's...
Amanda (37:25)
It's so bad, right? It was the worst we
found in our research.
Andy (37:51)
I, you know, Stephen Few 1500 words, slamming our definition of a dashboard in the big book of dashboards. You know, he said, these guys don't really have a right to teach about it. They can't even define it. I don't, give no shit to that at all. If you are building something that your users can use efficiently, then call it whatever the hell you want. I, you know, app. Yeah.
Fiona (38:15)
Do you know what though? That's
the second time that you've mentioned Stephen Few on the conversation today. I feel like maybe, no, no, second, second. Yep, yep.
Andy (38:22)
No, it's the first time today. It's the first time today. Is it? OK.
Sarah (38:24)
No, second, second.
Amanda (38:25)
Second, second, second, second. Yep.
Andy (38:29)
He got under my skin, man.
Fiona (38:30)
Yeah, I think there's something in this.
Amanda (38:34)
Do you know what though? You're not wrong, Andy. think that I also do not care one bit what people call it from a, what do we call it semantically and are you using the data? Great. I do think that when we have lots of different names for the same thing or don't have a shared definition when we say a word, that's when we end up getting in a lot of trouble, especially when you're working with clients who, when I say dashboard, they think one thing and I think something different.
Andy (38:34)
I-
Mm-hmm.
Mm-hmm.
Amanda (39:02)
I mean, I think if you say the word dashboard, you often think of something that's got lots of like filters and all those bits. And if you ask people to sketch it, it would all look somewhat similar. You'd have a multiple charts with some kind of interactivity or filters and have some way of engaging with the data. And I think data storytelling is equally fraught and plagued by like when we say a data story, some people immediately think it has to, because of data have charts with it. It doesn't.
those kind of pieces, but shared definitions help us when we're actually working in teams or working with clients. And so I do care what we call it within the scope of one project or one team so that when I'm talking to you, dear darling client over at John Hopkins University.
Andy (39:40)
Mm-hmm.
Amanda (39:47)
that when we're talking about a dashboard, we have the same vision-ish for what we're building. So we don't get to kind of the prototyping phase and I'm building wireframes that look one way and they're envisioning a very different kind of mobile data application that is very separate from what my expectations were. Now, I think that early stage engagement with users, prototyping, showing people things early on can help to resolve those definitional gaps. But I do think that within a team or within a project,
you do need to have a little bit of that level setting and defining of terms. But in the bigger sense of the world, I mean, think that a of people just like opining about why dashboards are dead. Jon Schwabish had one recently on the same thing.
Andy (40:24)
Yeah.
It's a really good thing for people to go and look at because it does give a lot more flesh on the bone. So Stephen Few had the super rigid, this is a dashboard, nobody shall disagree with me. We came along and went, it's this kind of vague thing. And then Nick Desbarats went, hey, everybody, let's stop fighting. Let's try and make taxonomy. So anybody who wants more detail, we've got more detail in the book. And Nick Desbarats has got a great framework as well to get some of those common definitions.
Amanda (40:53)
Nick loves those taxonomic
diagrams and decision charts more than any other data person I know. And I love them. They're so nerdy and I love them. I find great joy in reading through them and seeing his thinking.
Andy (40:56)
Yeah. But I...
I think my own journey, which I think probably a lot of data analysts go through, when you begin a journey of expertise learning, you know nothing. So you learn, wow, a dashboard is this. Brilliant. So then you know what a dashboard is. And then eventually, you know so much. You're like, but I've seen thousands of these things. I've seen thousands of use cases which break this definition. So you go from.
knowing nothing to thinking you know everything to where I am now going, I don't know. Dashboards look like Wordle. It could be a Wordle diagram. It could be a NASA big wall thing or a COVID dashboard. So that's how I ended up not reading. Well, it's not that I don't care.
Sarah (41:43)
I love your thoughts there, on making sure that people understand what the definition or are consistent on the definition of terms like dashboards. Because I've been into a client and thought I was going to build a dashboard and it was a 50 page PDF that I was building. That's a lot of disappointment in that day.
Andy (41:59)
You
Amanda (42:02)
That is very
different. I mean, I've also started out where someone had this idea that they wanted to build a dashboard. This is for a COVID survey project that I was working on. They wanted a dashboard and their vision was we have survey questions. We want people to find anything they want. So pick the dropdowns, all the filters, a lot of labor and to Andy's word earlier, friction on the user and they need to know what they're doing with it. And in our initial discovery work, we kind of defined out personas for who this was for.
And it was data across 67 countries on the adoption of different COVID prevention behaviors, knowledge about those behaviors, attitudes. And it was important data for health communication professionals and for policymakers to understand what was happening in these various different countries and how those knowledge, attitudes, and practices had changed. And what I think was interesting is we started out with this concept of this like, let's dive in and look at the data dashboard. But when you zoom out and say, this is for health communication professionals,
who are not data analysts, we need to serve up those insights in visually engaging ways with some micronography and other design touches. And the model we ended up using was more of a long scrolly dashboard that allowed people to go through and have lots of supporting text and pop-ups that helped them understand what the charts were telling them. And it was a very different vision of what that dashboard would be, but it was ultimately much more effective and reusable for slides and for presentations.
then a kind of serve it up yourself kind of dashboard would have been with, mean, Andy's, what was it? 10 filters, 25 filters? How many filters were on your first Oxford dashboard? Having that many filters across the top and expecting your users to carry the burden of knowing what they would be able to find versus serving up information in a more consumable, I think one of you said the tik-tokification, but a more bite-sized way of consuming information.
Andy (43:32)
Many. Yeah. β
Yeah.
Sarah (43:53)
Yeah.
Fiona (43:54)
Yeah, for sure. mean, I love where this conversation is going around friction and removing that friction and really helping people like meet people where their data needs actually are. Andy, you know, you've talked about, let's just forget about the definition of dashboards or the word dashboard and figure out what is the definition behind it. I think the reason why we talk about that
a lot is so many people are afraid of losing their jobs to AI or, you know, worrying about job turnover. And we're seeing it happen a lot in market as well. If you could sit down with a chief data officer or an analytics leader for 30 minutes, what's the one thing you'd want them to understand about how their teams are working with data right now or how they should be working with data?
Andy (44:26)
Mm-hmm.
God, that's a difficult question. That's a great question. I would say that your team are probably using AI extensively and not telling you. And that's probably because 18 months ago, a chief ex-officer, you said, don't use AI. And they all did anyway, right? So I think the leadership created a problem in a lot of organisations in that way. There's nobody
Amanda (44:47)
Such a good question.
Sarah (44:49)
You
Andy (45:10)
Nobody is an expert in all of the AI technologies. It is so disruptive, and it is moving so terrifyingly quickly that you just can't keep up. I'm trying to do an AI for data analysts course at the moment for one of the big training providers. I started writing my course draft about three weeks ago, and I've pretty much had to redo a whole bunch of it because literally, it's changing. It's really hard. We wrote a chapter in the book about AI.
So it's all happening. People are using it. People are being incredibly productive with it. I have a pit in my stomach almost constantly when I'm thinking about AI's impact, because on the one hand, I can see it making people more productive. And leadership in companies have responsibility to drive
drive value for their company. They can do that by being more productive or they can save costs. Right now, across the industry, across the tech landscape, CEOs are laying people off because they're being, well, because they are saying they're being more productive, but they're just saving costs. And I've spoken to CEOs who have told me that that is their goal with AI and I hate it. And so the pit in my stomach is because
The technology is amazing. Just the things that Claude has added in the last few weeks are breathtaking. I've built applications in Claude that I could never, ever, ever have done. And yet I see this coming storm that I don't know what to do about. So I don't know what I'd say to a CDO because I just have a pit in my stomach that is because I'm excited by the technology, but I absolutely loathe the theft of IP.
lack of regulation, the environmental destruction, the exploitation of labour and what it means for capitalism. There you go.
Sarah (47:02)
It's a big answer. Over to
Amanda (47:04)
I
Sarah (47:04)
you, Amanda.
Amanda (47:05)
love when you end at a high point, Andy, and really close it out with a really positive outlook for where things are going.
Andy (47:09)
Well,
yeah, I had a keynote on this and someone said, Andy, you should do the negative bit first and end on the high. I'm like, yeah, okay. So I did that.
Sarah (47:17)
You
Andy (47:20)
Yeah. I might have talked about this last time, but this book completely shook, has shaken a lot of my world foundations. It's all about Sam Altman and open AI, and it's absolutely amazing. But read it and it will enlighten you. yeah, sorry. Empire of AI by Karen Howe. It's fantastic.
Amanda (47:32)
It is. It's a very good audio book too for anyone who is an audio listener here on this listening to a book and it was great.
Sarah (47:33)
And for those of those listening online, that was the empire.
Thank you. And on that bombshell that Andy's dropped, Amanda, would you like to answer that question?
Amanda (47:41)
And that's it. And we'll see you next time.
Andy (47:44)
lol
Amanda (47:47)
I have 30 minutes in a room with a CEO or CDO.
Sarah (47:48)
has a mentor for.
Amanda (47:52)
To me, it would depend so much on the industry and what I would want to say to somebody. But I would say that as people are exploring and looking at how they're using data right now, I would...
Andy (47:55)
Mm.
Amanda (48:03)
look at the ways in which people are being creative and thoughtful about making data more accessible, but in low friction, lower technology ways. I would tell a CEO to double down on creating space for in-person communication and connection between staff and other folks. Because at the end of the day, if you start leaning down your team such that you only have a couple different data analysts sitting on a team who are primarily collaborating with another technology tool,
doing all of their work and their iterations and their feedback loops and everything else, you're going to eventually lose people who leave your company voluntarily because they don't enjoy their job anymore. A lot of us went into the data visualisation domain in some capacity because we like doing this kind of work. I want to find tools that make the sludgy stuff that I don't like as much easier to do so I can do the more complicated, interesting, engaging with other people parts of it. And so I would really...
Andy (48:41)
Mmm.
Amanda (48:58)
probably avoid all the conversations about predicting the future of technology. And instead, try to remind people to think about the fact that when you're leaning out your teams a lot for all of those cost savings, if you get down to a place where people are collaborating maybe with, mainly with robots, that there will probably be some negative implications from a data side, but you'll also eventually lose those really good people because a lot of the joy we find is in that human connection.
Fiona (49:21)
Aww.
Andy (49:22)
I I'd
have saved money.
Amanda (49:23)
But they'll have saved money, capitalism. I'm sorry, I'm sorry. I'm based in the United States. I should be celebrating and championing the shareholder value. They may have added shareholder value. And I know that that is the ultimate KPI here in the U.S.
Andy (49:26)
Sorry. Yeah.
Ha
Fiona (49:40)
globally.
Sarah (49:40)
everywhere.
Andy (49:40)
the
UK. Yeah everywhere.
Fiona (49:42)
for sure. Okay, so before we get into the quick fire round, I would like to wrap up with one longer question, which is really speaking to our listeners out there. You know, if an analyst came to you and said, nobody listens to my work or nobody uses my work, what's your first question back at them?
Andy (50:04)
Did you ask them what they wanted before you made the work?
Amanda (50:08)
I would ask them to tell me more about who they were sharing work with and if they had a vested interest in what they were sharing and an expectation of what they'd do with that work. I also would ask, are you someone that they know and trust?
Fiona (50:23)
That's a tough one.
Sarah (50:23)
It's a good one.
Andy (50:24)
Yeah.
Fiona (50:25)
β
Sarah (50:25)
Because I analysts
like to sit in the background and, you know, don't want to be known and yeah, interesting.
Amanda (50:33)
But I think that's
why a lot of the people I've seen be really successful as data analysts, especially where they're thinking about how to answer business questions with data, are people who had interesting early in life journeys through many different service industries. I I worked as a nanny, a lifeguard, a snack bar attendant, like in retail for a while, and did all of those servicey jobs before I ever became a consultant. And I worry a bit that in a world where
Everyone's so pressured to do high pressure tech internships and be in company environments that all of those kind of interpersonal interaction skills and ways in which you build rapport with people so they do trust you gets lost. Cause we learn a lot of those things through those. I mean, I struggle to say early career, but kind of early career moments that you have in high school and college and those early jobs where you learn to understand and listen to people.
and then respond accordingly with the information you're giving or the services you're providing.
Andy (51:30)
I can't believe this book is 10 years old, Fuzzy Techie by Scott Hartley. This is great. This book was all about, you know, the people who are gonna work in the AI driven world are the ones who did humanities degrees, liberal arts, because they are trained more to be empathetic, to listen, and to do critical thinking and project handling. So I think that alludes to what Amanda was saying.
Amanda (51:54)
What are all of your undergrad degrees in? Now I'm so curious.
Fiona (51:54)
Yeah.
Andy (51:58)
Geography, liberal arts. So I was, this was confirmation bias for me. Yeah.
Fiona (52:04)
at its finest.
Amanda (52:04)
That's why I held onto
it.
Fiona (52:06)
So Amanda, you mentioned building trust. Have you got a quick tip on how you can suggest that these people build trust?
Amanda (52:12)
how quick tips on trust building seems like a fool's errand. That trust is something that happens over time. think that my big tip to people would be to be someone who follows through when you say you're going to do something, do it. And I think that that's actually really undervalued and something that doesn't happen all the time these days. But even on small things, you say you're going to send the follow-up email, you say you're going to go pull the number, do it and do it in a timely manner. And make sure that you're doing that consistently and that the work that you're producing and sharing, you are...
providing and sharing consistently so that you go ahead and you build that rapport over time so that when an error happens, because it will, whether you're using AI tools or not, you have that rapport built up with your stakeholder that they know you, they trust you. And you can hopefully, when that error happens, you can dig into the details and explain to them what happened and how you're going to prevent it from happening again. Because I think that process loop is really important when errors happen to make sure that someone continues to trust that you're not just someone who can deliver good work.
but can continue to improve your work over time.
Fiona (53:10)
Awesome. Thank you.
Sarah (53:12)
All right, we are ready for the quick fire round and we want instinct short answers. Now just, I'm just going to say that I both Fi and Andy are really bad at this. So, β okay. So short answers, no caveats. This is that quick fire and we're asking both of you. So even if you disagree, it's even better. Let's go.
Amanda (53:24)
I am too. Have you heard me talking today? I'm terrible at this.
Fiona (53:28)
it.
pie charts, defend them or abolish them? Amanda?
Andy (53:43)
friend. sorry.
Amanda (53:43)
Defend them.
Fiona (53:45)
Okay, cool.
Sarah (53:46)
Dashboards in 10 years, more or fewer than today.
Amanda (53:50)
Fewer. Different.
Andy (53:52)
you up.
But still buy the book though.
Fiona (53:53)
AI did.
Amanda (53:55)
and still valuable. The ones that survive the culling are going to be good.
Fiona (53:58)
Not Kevin!
Sarah (53:58)
I'm like, stop
it!
Fiona (54:00)
generated insight or human curated story.
Andy (54:03)
human.
Amanda (54:03)
Human
curated story.
Sarah (54:04)
The most dangerous phrase in data communication.
Andy (54:08)
sorry, it looked bigger on my laptop.
I will walk out of your talk if you say that in a presentation.
Amanda (54:15)
This chart tells the whole story.
Andy (54:17)
that's how we should, but there's a whole series of those lines. my God. I can't cope.
Fiona (54:22)
One thing every data team should stop doing immediately.
Amanda (54:26)
making their work invisible, promising one big deliverable, handing it off and never showing the details that got you there.
Andy (54:32)
Saying you can't download it to Excel
Fiona (54:34)
That's a good one too. And just on making it visible, we can do a short plug here. Sarah and I are actually speaking at Tableau Conference this year on how to become a data fluencer. So that's all about making your work really visible and amplifying it across the organisation. So if you're coming along to TC, we'd love to have you. That was not prepped, but I thought I'd take the opportunity to sell, sell, sell. All right.
Andy (54:57)
sell sell sell
Sarah (54:59)
Hahaha
The best dashboard you've ever seen was...
Amanda (55:03)
best.
My weather app, the YR Weather app. I use it every day. I love it.
Fiona (55:07)
Nice.
Talked about great legs. Yes, how many times have you spoken about it?
Sarah (55:10)
Hahaha
Amanda (55:11)
You've got two whole books.
Andy (55:12)
What, Iraq bloody toll? Yeah,
but the best I've ever seen. The relation between the smallest thing that's given me my biggest career boost is that little tiny chart. My career can be condensed down to turning a bar chart upside down, but it wasn't a dashboard.
Fiona (55:18)
The data story?
Amanda (55:32)
So there's before Iraq's
bloody toll and after. Can we turn that into an acronym for Andy?
Fiona (55:37)
you
Andy (55:38)
Yeah, I don't know. There you go.
God. What's my favourite dashboard? I don't know. My favourite... The one I'm... My favourite from the book is the NASA one, right? So I'm going to go with that because the success metric, as Amanda said earlier in the episode, was not that people understand the minute differences between this climate measure and that climate measure.
but they got a sense of the data and it was perfectly designed for the environment in which the people see it. And I think that is a brilliant, brilliant use case.
Sarah (56:11)
What's the one thing you both want our listeners to take away from today's conversation? Start with you, Amanda.
Amanda (56:17)
Even if you build different dashboards and data visualisations, that doesn't mean they're going to get used. You have to actually spend time thinking about who the audience is, what they need to know, and making sure people have the right data at the right place at the right time. And that is valuable human insight and context that I still think dashboard developers, designers, analysts, business analysts, all bring to the table. So even in a world obsessed with AI,
you are still bringing valuable context and insight to the table and are doing valuable work.
Andy (56:47)
Mine's fairly similar. think the reason I love this field is because there is a unsolvable tension or infinite flexibility about the ways in which you can communicate data. Even a simple bar chart can be framed completely differently, as you have seen with the Iraq's Bloody Toll many, many times. And that is the marvelous creative power that data analysts have. Whether you want to come in
as a science nerdy engineer or come in as a artist data communicator. There are so many paths into this field and it creates a and should continue to create a diverse, rich profession of empowered people.
Fiona (57:27)
This PODDY has been absolutely incredible and you both have been so generous with great suggestions and ideas that I'm sure that our listeners can take away and start to apply immediately. Amanda, Andy, thank you so much for being amazing guests on UnDUBBED
Andy (57:48)
Thank you so much for having us.
Sarah (57:49)
Yeah, and what?
Amanda (57:49)
Thank you for hosting
us.
Sarah (57:50)
We're going to have, I think, like a whole scroll of links in the show notes. But for those that are listening today, where can people most reach out to you?
Andy (57:55)
Yeah.
I'd say my site is howtospeakdata.com. That's got a link to my sub stack newsletter and I'm most active on LinkedIn.
Amanda (58:09)
You can find me on LinkedIn under my name. Also, you can find my website, amandamculloch.com, which has links to my sub stack, Vis Responsibly, as well as other resources on working with me. And you can find both of us on dashboards.deliver.com along with a free three chapter preview of the book. So if you're curious about the book, about the framework, anything else, we have a lot of great resources there, including a dedicated notebook LM environment where you can explore themes from the book.
or find downloadable resources under our Dashboards fan club dropdown. So we'd love to have you connect with the content there.
Fiona (58:43)
definitely want to go and check that out. was something that I did not know. So thanks for sharing that. And our listeners, if you love this episode, make sure you hit follow, leave us a review and share it. Please share it with anyone who's ever built a dashboard that nobody used.
Amanda (58:45)
You
Sarah (59:00)
And I'm just going to re-quote Andy. I wrote down a little quote that he said before.
Andy (59:04)
the word dashboard is just an example of semantic drift because it used to be something at the front of a stagecoach, a board of wood that stopped water dashing onto the driver's boot.
Fiona (59:15)
That's so poetic.
Sarah (59:16)
So poetic.
Amanda (59:17)
So low tech team, so low tech, yet here we are still calling something a dashboard centuries later.
Sarah (59:19)
You
Andy (59:21)
Yep. Yeah.
Fiona (59:24)
So what I'm taking away is the future of data isn't more data, it's better answers. Bye.
Sarah (59:29)
See you next time on UnDUBBED
Andy (59:30)
I
