YT YouTube2Text

YouTube2Text → Transcripts → storytelling with data: #1 the art of feedback

storytelling with data: #1 the art of feedback — Transcript

by storytelling with data · 5,825 words · 834 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00[Music]
0:03welcome to storytelling with data the
0:06podcast where listeners around the world
0:08learn to be better storytellers and
0:10presenters with bestselling author
0:12speaker and workshop Guru Cole nisb
0:15naick we'll cover a wide range of topics
0:17that will help you effectively show and
0:19tell your data stories so get ready to
0:22separate yourself from the mess of 3D
0:25exploding pie charts and deliver
0:26knockout presentations and with that
0:29here's here's Cole hi I'm Cole and I
0:32focus on telling stories with data today
0:36you are listening to my very first
0:38podcast I'm guessing if you're hear it
0:40means you might be familiar with some of
0:42my other work perhaps you follow the
0:44blog or have read my book the podcast is
0:47going to be a way for me to really put a
0:49voice behind the data and concepts for
0:52communicating with data the focus of our
0:55podcast today is going to be on giving
0:57feedback through this you'll get a sense
0:59of some of the things that I've been up
1:01to recently uh on the blog discussions
1:04I've had at workshops and behind the
1:06scenes we'll also tackle reader Q&A
1:09where we'll talk about when and why we
1:11should use graphs in the first place
1:14considerations for dashboards and
1:16Reporting and I'll also outline some
1:18data viz 101
1:20resources when it comes to this podcast
1:23I'm really trying this out I want to see
1:24what sort of interest level there is for
1:26this medium and the type of content that
1:29I want to talk about about and I'll
1:31definitely be iterating over time and
1:33welcome your feedback on that front and
1:35really any sort of feedback here will be
1:38useful everything from this is great
1:40keep doing it to hey that was awkward
1:43maybe rethink that next time or here's
1:45some content that I'd love to see in a
1:47future episode uh all types of feedback
1:50would be great and you can provide that
1:52by sending oute to feedback at
1:54storytelling withth
1:56data.com speaking of feedback giving
1:59feedback is is something that's come up
2:01for me in a lot of different ways lately
2:05uh I think there is immense value in
2:07being able to both give and receive
2:11feedback when it comes to refining our
2:13data visualizations and the way that we
2:15communicate with data broadly and so
2:18today I want to talk about some
2:20considerations when it comes to giving
2:22feedback on data visualizations and I'll
2:24do this through a few different examples
2:26and scenarios that I've encountered
2:28lately first off I want to talk about
2:30the economist hurricane challenge uh
2:34this was one that I ran on the blog back
2:36in September and actually early on in
2:39September I found myself riding the
2:41subway in New York City was there for a
2:44couple weeks doing some work and as I'm
2:46riding the subway I'm scrolling through
2:48my feed on feedle and I come across one
2:52of the economist daily charts and it was
2:54titled hurricanes in America have become
2:57less frequent and I spent some time
2:59staring ing at this graph realizing I
3:02had some gripes with it uh but not
3:04really the time to do anything about it
3:06to fully critique it or do a makeover at
3:08that point so I thought well I'll turn
3:10it over to you my readers uh and have
3:13you share your views and makeovers now
3:16let me back up and explain a bit about
3:19what the graph looked like so this was a
3:21stacked bar chart the y axis the total
3:25uh showed the number of hurricanes that
3:28had hit the US
3:30and it showed this over time which was
3:32the xais component and then each bar was
3:35split up by the category of hurricane 1
3:38through
3:39five and the response that I received to
3:42this was much more than I anticipated so
3:45I received 60 makeovers uh with readers
3:49raising a ton of what they believed were
3:52issues with the original graph and I
3:55thought it was interesting to see the
3:57tone of many of these notes that came to
3:59me
4:00uh you know the economist did this wrong
4:02or they made that bad decision and some
4:05people were pretty brutal which it's
4:08actually incredibly easy to be when
4:11you're at arms length like this it's
4:13easy to forget that there is a person or
4:16a team who made the various design
4:19choices for a variety of reasons and
4:23actually didn't think about this so
4:24directly at the time but was reminded of
4:27it very clearly when the economists
4:29emailed me so Alex Selby boooy who is
4:33head of data journalism at The Economist
4:35wrote me dear Cole thank you for
4:38organizing your fascinating how you
4:40would visualize Hurricane's challenge
4:42here in The Economist data journalism
4:44Department we've poured over every entry
4:46with the same degree of interest and
4:47healthy debate that our original chart
4:49seems to have provoked as with all data
4:52visualizations we had to make some
4:54compromises the incomplete final decade
4:57was as many people noted unsatisfactory
5:00instead of leaving it to our readers to
5:01correct for the shorter time period in
5:03hindsight we could have used five-year
5:04groupings like Robert did or 15-year
5:06clusters like Todd did to maintain
5:08consistent spacing on the x-axis we
5:11could of course have improved other
5:12elements as well with more time we could
5:15have Incorporated every 2017 hurricane
5:17up to and including hurricane at Irma
5:19while accounting for the fact that this
5:21year's season is not yet over as Sharon
5:23noted we could have considered hurricane
5:25damage too or looked Instead at
5:27accumulated Cyclone energy that said
5:30these data cover all Atlantic storms and
5:33we wanted to focus just on those that
5:34made landfall in the United States as
5:36our article was pegged Irma hitting
5:38Florida the previous day as a
5:40description of what has occurred in the
5:42past we thought that the headline and
5:44subheadline on the story hurricanes in
5:46America have become less frequent but
5:48the most damaging storms now appear
5:50appear to be slightly more common was
5:52accurate and we felt that the trend
5:54lines helped show this but it's
5:56interesting to note that for every
5:58visualizer who us used trend lines to
6:00reach a similar conclusion there was
6:02another who loudly dismissed them as
6:04misleading or is equally vocal about
6:06something else we had done wrong it is
6:07heartening to be part of a data
6:09visualization community that has such a
6:11strong voice all the best do let me know
6:13if we can be involved in any future
6:15challenges so I thought it was awesome
6:18to see this note and the attention from
6:21The Economist Dana journalism uh
6:23department and it's a pretty amazing
6:27balanced response when you consider it
6:30that they could have come from a very
6:32defensive position given that we have
6:34these 60 makeovers of different people
6:36outlining different issues or potential
6:39issues with that and yet the way I read
6:42it at least it came across as a really
6:44balanced response and I also thought it
6:48was great the way that Alex uh
6:50highlighted the constraints right the
6:52the fact that they made compromises and
6:54tradeoffs uh for a variety of reasons
6:57that's one of those things that when we
6:59remake data and redo data visualizations
7:03we just we have no visibility into those
7:06constraints that the person originally
7:08visualizing the data faced uh and I was
7:12reminded of this same idea recently I
7:14was following a heated Twitter exchange
7:17uh between a few folks uh Alberto Cairo
7:20Elijah Meeks and Steven Redmond by the
7:23way one of the things I think is just
7:24fascinating about Twitter so Alberto's
7:27in Miami uh Steven's in
7:30Elijah is in California right I'm here
7:32following from California so just the
7:35way that applications like Twitter allow
7:38these conversations to happen in a way
7:41that would otherwise never be possible
7:43but so they they're having this debate
7:45and uh Alberto links to an article at
7:48one point the article is one back from
7:502015 by Fernand viegas and Martin
7:53wenberg the article is called design and
7:56redesign in data visualization and in
7:59this article they talk about this idea
8:01of popular criticism and how the data
8:04visualization field has become ripe for
8:07drawing this popular criticism in a way
8:11that you really can't do for other
8:13mediums right if you think of a book or
8:17a movie someone can critique those
8:19things but they can't actually remake
8:22them from start to finish right if
8:23you're critiquing a book maybe you
8:25rewrite a sentence but there's no way
8:26you're going to go in and rewrite the
8:28book whereas in data visualization this
8:31actually happens right if you can get to
8:33the underlying data set anybody can
8:36remake a graph and so in many ways this
8:40is good it's a way that people can learn
8:44but it can also be problematic and
8:47that's what the article really focuses
8:49on and some of the ways that this can be
8:51uh problematic so one of the things they
8:53highlight is how hindsight can be 2020
8:56and to illustrate this they look at
8:58Edward tuft's criticism and remake of
9:01data from the Challenger space shuttle
9:04disaster so it's it's easy knowing what
9:08the important things were uh in that
9:11disaster to now go back and make data
9:12visualizations that make that brutally
9:15clear whereas at the time that wouldn't
9:17have been possible in the same way
9:19because that information simply wasn't
9:20known at that point now another way that
9:24this popular criticism can be
9:27problematic is in the removal of context
9:31and this really reminds me of the
9:33hurricane makeover and that feedback
9:35that we heard from Alex uh that those
9:38who are doing the makeovers and
9:39providing feedback they just have no
9:42visibility into the constraints that
9:44those visualizing the data faced and
9:48Fernanda and Martin write design is
9:50compromise they also outlined some
9:52general guidelines simple guidelines for
9:54data visualization critiques one
9:57maintain rigor second secondly respect
10:00the designer and thirdly respect the
10:03critic and I think tone and framing
10:06become so important when we're giving
10:09feedback on data visualizations so that
10:12this is coming from a place of helping
10:14or friendly debate and not um not
10:18criticism in in the sort of negative
10:21aspects of that word in any case
10:23excellent article I'll make sure that we
10:24link to it and I consider it a mustre
10:26for anyone who is creating or critiquing
10:30data
10:31visualizations I actually had a recent
10:34request for critiquing a data
10:36visualization uh this came from a
10:38follower who's been working in a company
10:41where he says historically we've had
10:43this really data heavy presentation done
10:45by our marketing team he shared the the
10:47presentation with me and it had uh a lot
10:49of dnut charts a lot of color and he
10:51says I recently hired someone though to
10:53specialize in data visualization and I
10:56bought him your book and so now this new
10:58data visualization Guru has revamped the
11:01ongoing version of the presentation so
11:03stripped out the Clutter made some other
11:06changes based on lessons from the book
11:09and he says we just delivered the
11:11presentation for the first time and we
11:13were met with mixed feedback uh but we
11:15want to make sure what we're doing is
11:17grounded in best practices and not just
11:20based on someone's personal preference
11:21and so he asked me to weigh in on this
11:24and so this scenario it would have been
11:26really easy to jump to something like
11:29you here's all the reasons that you're
11:31right as I see you applying Concepts
11:33from my book and here's all the reasons
11:35that they were wrong you know with their
11:36donuts and their color but I actually
11:40don't believe that's necessarily the
11:42case uh and I definitely don't believe
11:44that's what would have been most helpful
11:46for anyone in this scenario so actually
11:48in this case I brought it back to
11:50audience if your audience doesn't like
11:53your approach no amount of telling them
11:56why they should like it or why they're
11:58wrong for not liking it is going to help
12:03rather for me this means we're missing
12:05something that our audience needs and
12:09so going back to I mean the the
12:13different things that we think about
12:14when we're crafting visualizations and
12:15we're following best practices and I
12:18actually wrote uh what some of this
12:19looks like from my perspective in a blog
12:21post back in September uh my guiding
12:23principles and outlined the different
12:26things that I'm thinking about when I am
12:28creating
12:29uh and critiquing data visualizations
12:32and the last guiding principle the final
12:33one is that audience trumps all else I
12:38think to be successful when it comes to
12:40visualizing data in a business setting
12:42we really need to be thinking about our
12:44audience first and foremost and if
12:47there's something that they are confused
12:50by or resist then we need to revisit the
12:55design and revisit the way that we're
12:57doing it and and think about what our
12:58audience needs are and how we can make
13:01sure those needs are being met so I
13:03think when we can do that then we lower
13:06the chances of of having that sort of
13:09resistance now when it comes to giving
13:11more specific data visualization
13:14feedback I encounter this scenario in
13:17the workshops that I do regularly so
13:20when I go in to do a half day or full
13:23day uh workshop with a team I'll I'll
13:25typically solicit examples from the team
13:28ahead of time and after we go through
13:30the core lessons then we look at some of
13:32the group's specific work and we discuss
13:36how we can apply the lessons that have
13:37been covered to uh the folks specific
13:41work their work and their colleagues
13:43work now as you can perhaps imagine this
13:46can be a very sensitive topic right this
13:49is someone's work that now a room full
13:52of people are critiquing and this is in
13:55the absence of those constraints as
13:57we've talked about and without actually
14:00having to do any work right they're not
14:02even remaking the examples they're just
14:04outlining some gripes and maybe making
14:07some sketches now the way I frame this
14:10up is it's not an exercise in ripping
14:13anything apart but rather thinking about
14:16how we can discuss and how we can apply
14:20the lessons that we've covered to make
14:23the selected examples even stronger and
14:26I tell people you now have dis
14:29Discerning new eyes when it comes to
14:31looking at graphs and this tends to be
14:33both a blessing and a curse uh but I
14:36challenge folks to use those new eyes
14:39that they have for graphs not only to
14:41critique because as we've seen that's
14:44easy to do uh when we're looking at
14:47other people's work but also critiquing
14:49right applying those same new eyes to
14:52their own work because I think by being
14:55harsh critics of our own data
14:58visualizations we really put ourselves
15:00in a place where we can continue to
15:02refine them uh and ultimately make them
15:05successful now that said doing this sort
15:08of critique and especially receiving
15:10feedback can be hard we get attached to
15:14our work there was a recent article from
15:17ux planet on medium called the subtle
15:19art that differentiates good designers
15:22from great designers in the article they
15:24list five differences between good and
15:26great designers and the first one that
15:28they list really resonated with me they
15:31said for great designers a critique is
15:33just a critique not an insult and this
15:37is one of those things that's really
15:39easy to say and incredibly hard to
15:41actually uh put into practice we get
15:44attached to our work and as we do we
15:47lose the ability to objectively see
15:52right when I've have put together data
15:54visualization I know where I want my
15:55audience to look I know the things that
15:58I want them to connect I know what I
16:00want them to walk away with uh the
16:03challenge though is making all of that
16:05tacit stuff in my head visually clear to
16:08my audience so that they look at my
16:11graph they see it in the way that I want
16:13them to now one of the ways that I get
16:16this sort of feedback is from readers
16:18where I'll post something on the blog
16:21and then I'll have people who volunteer
16:23makeovers of my makeover or provide
16:25feedback and now in some cases this can
16:27be challenging right cuz as we've talked
16:29about those doing makers overs they
16:31don't necessarily have any visibility
16:33into the constraints that we face um or
16:36for me you know the specific lesson
16:37maybe that I was trying to highlight
16:39through a particular
16:41makeover um but this is also an
16:43incredibly helpful part of the process
16:47uh both for me as well as a person
16:49providing the critique because it forces
16:51us when we critique to articulate our
16:53logic in a way that not only makes sense
16:56to us but that will make sense to
16:57someone else which can help us think
16:59through things uh in new ways um this
17:02actually happened recently uh this uh
17:05reader feedback uh in a blog post or
17:07after a blog post that I had posted um
17:10this was back a couple weeks ago the
17:11blog post was titled A Tale about
17:14opportunity and uh in in the blog post I
17:17was looking at an example from the
17:19pharmaceutical industry where the
17:21original graph um was a 100% stacked bar
17:25chart where the focus was uh
17:29the the bar on the left was the number
17:31of patients who had been diagnosed with
17:33a particular illness and then the bar on
17:35the right represented the patients
17:36taking uh this drug for the particular
17:39illness and then the illness itself
17:41could be divided into different
17:42severities mild moderate severe and this
17:46was a case where we could actually
17:48switch the way that we were categorizing
17:51uh and instead of making the primary
17:53categorization those diagnosed versus
17:55those taking and then the categorization
17:58within that the severity of the disease
18:00but actually breaking it out by the
18:02severity of the disease first and
18:04foremost and then within each of those
18:06stages looking at the number of people
18:09taking the drug out of the total
18:11diagnosed meant that we could flip it
18:13around and really highlight the
18:15opportunity so doing it this way allowed
18:18us to see not only who's taking the drug
18:20but who else could be taking the drug
18:23and to visualize this opportunity I
18:26showed empty space that was outlined and
18:29so imagine a horizontal stacked bar
18:32chart where the bottom series is dark
18:34blue this represents the number of
18:37patients currently taking the drug and
18:39then stacked on top of that is a white
18:42series that's outlined in that same dark
18:44blue which represents the portion who've
18:46been diagnosed with the dis disease but
18:48aren't currently taking the drug so
18:50those who could be taking the drug then
18:52in the final version I had some
18:54annotation on there listed out some
18:56percents um in that same dark blue and
19:01the percents Were Meant to describe the
19:03white portion of the bar that was
19:04outlined in dark blue but it was
19:07probably 20 minutes after I posted the
19:09blog post that I had a couple of reader
19:11comments saying hey that's confusing
19:13because when I read it and I think about
19:15it I can figure out that I meant to tie
19:18those percents to the white portions but
19:19when I first look at it I want to tie
19:21the blue numbers to the blue portions of
19:23the bars this is one of those things
19:24that once somebody points it out yes of
19:26course that was a silly thing to do on
19:28my part of course it's natural people
19:30are going to want to connect those
19:31things that look similar and I didn't
19:33even see it though when I was doing the
19:36graph because I knew where I wanted my
19:37audience to connect those pieces and so
19:40in this case I used that reader feedback
19:43to actually go back and refine my final
19:46Visual and added another one at the end
19:49that showed the opportunity outlined in
19:51black and change the percents that were
19:53meant to describe that opportunity to be
19:56in bold black text so this this was a
19:58case where I was able to use some reader
20:01feedback to make the example even more
20:04effective and now for me it's rarely
20:07about this being better than that when
20:10we offer critique or provide feedback uh
20:14but rather more about optimizing
20:16trade-offs because is we've discussed
20:18pretty much every decision we make when
20:21we're designing data visualizations and
20:23the slides and pages that contain them
20:26these decisions involve trade-offs so
20:28for me it's a without understanding what
20:29those tradeoffs are and making smart
20:33decisions in light of those things so
20:36ins some giving feedback is a hugely
20:39important process of helping everyone be
20:43better when it comes to visualizing data
20:45and I I do believe that when we provide
20:48feedback we should be respectful and
20:50consider tone and framing right as we've
20:53discussed we're critiquing someone
20:55else's work and that person likely faced
20:58constraints that we have absolutely no
21:01visibility to but I think also as
21:04designers of data visualization we need
21:06to have thick enough skin to recognize
21:09critique and take it as such take it as
21:12critique not as an insult so I'd love to
21:15hear your thoughts on this topic uh
21:18please comment on the blog post and we
21:20can continue the discussion
21:22there let's shift gears now though and
21:25spend some time on reader questions so I
21:28announced this podcast through a blog
21:30post a few weeks ago where I talked
21:33about the fact that my kids ask a ton of
21:36questions and that I've come to
21:38recognize over time that this is a
21:40really important part of their learning
21:42process and was recognizing sort of
21:44simultaneously that I receive a ton of
21:46questions from readers and those who
21:49follow my work as well so I thought a
21:51great segment in the podcast would be
21:53take some time to answer some of those
21:55questions so uh want to answer some of
21:57the questions we've received to date
21:59also if you have a related question that
22:02you would like to pose you can send that
22:04to ask Cole at storytelling withth
22:08data.com Matt says my question is very
22:10basic essentially when should I even be
22:14creating a chart as opposed to just
22:16relaying my big idea through text I work
22:19in digital advertising and when I'm
22:21delivering insights to clients it's
22:23usually text in one or a couple of
22:25sentences I can understand when a chart
22:27is useful if if you're dealing with a
22:29large number of variables and you just
22:30want to focus on one thing but for these
22:32relatively simple insights would you
22:34even recommend making a chart it's
22:36fascinating to learn the art of
22:37beautiful and useful charts and I just
22:38want to make sure they keep being useful
22:41to the audience so this is a great
22:43question and I love the thoughtfulness
22:45behind it and I think anytime we think
22:47about showing data we should ask
22:49ourselves the question of do we need to
22:51show this
22:53data just because we have data doesn't
22:56mean we necessarily need to show it and
22:58I think for for me the sort of things
23:00that I want to consider when it comes to
23:03whether I should be showing data uh one
23:05is just being aware of the value of
23:09visuals for making data accessible so
23:12helping someone understand or see or
23:16remember something uh images are
23:19powerful graph being one sort of image
23:21in their ability to prompt recall so if
23:25there's something in the shape of the
23:27data or there's something about showing
23:30it that's going to help you reinforce
23:32the message that you want to make I
23:35think there can be immense value in
23:37pairing the visual the graph with the
23:40verbal the story so when I do that now
23:42not only can my audience remember what
23:44they heard but they can also remember
23:47what they saw I think another important
23:49component is certainly your audience is
23:53your audience going to need data in
23:55order to trust you or be convinced
23:59uh if so then you want to think about
24:00how you can weave that data directly
24:04into the story that you want to tell so
24:06I think ultimately graphs uh use them
24:09when they can help someone see or
24:11understand or remember and when they're
24:14going to be useful for your
24:17audience Brian says my question for you
24:20is how can I can apply your principles
24:22and methods when creating and designing
24:23dashboards I work for a large Hospital
24:26in Boston and lately most of the request
24:28I get are to create dashboards the issue
24:31is that the users consuming them want
24:33everything on it I'm struggling with
24:35this because I know from experience in
24:36following your principles that a lot of
24:38the information they want is unnecessary
24:40please help so I tend to draw one uh
24:44important distinction which is the
24:45distinction between
24:47exploratory right what you do to
24:49understand the data when you look at it
24:50this way and that way uh to find the
24:53interesting insights and explanatory
24:56which is once you found the interesting
24:57insights and you want to now communicate
24:59those to someone else now for me
25:02dashboards any sort of regular reporting
25:04these fall more into the exploratory
25:07space where dashboards can be super
25:10useful at putting a lot of data into a
25:12small space where I can scan through and
25:14I can look to see where are things in
25:17line with my expectations or where are
25:19they not in line with my expectations
25:21where might there be interesting things
25:23that I can further dig into to get at
25:26the story now one thing to keep in mind
25:29is just because the dashboard is how you
25:31find the interesting takeaway or story
25:34does not necessarily and I would say in
25:36many cases does not mean that the
25:38dashboard is going to be the best way to
25:40then tell that story to your audience so
25:44I think the dashboard can be incredibly
25:46useful for getting at that story and
25:48then when you have something specific
25:50you need your audience to know or an
25:52action you want them to take then you
25:53actually want to take that data out of
25:55the dashboard and do a lot of things
25:57that I talk about in the book and on my
25:59blog to make that story both visually
26:02and verbally clear now when it comes to
26:05dashboards and designing effective
26:07dashboards one resource I'll recommend
26:09is a book that came out earlier this
26:10year called The Big Book of dashboards
26:13and this is a collection of um about 30
26:14different dashboards across various
26:16Industries um so it can be nice just to
26:18flip through and see ideas for how
26:21people have visualized different uh data
26:24and then the actual content is the
26:26author's commentary about what works
26:28well and what they might have done
26:30differently had they been designing the
26:32dashboards and goes through some
26:33alternative um Al or alternate uh
26:36visuals there so it's a nice thing to be
26:39able to flip through to get ideas but
26:40then also um some great uh great uh
26:45content to help develop understanding of
26:47best
26:48practices Nuru says my question would be
26:51how to get people in the organization to
26:53think more analytically and not generate
26:56reports for the sake of generating
26:57reports me the IT team doing data
27:00analysis which can be a bit tricky since
27:02the business teams are used to having it
27:03do a quick fix when something is broken
27:06and not so much for process Improvement
27:08so this actually makes me think back to
27:10my time at Google and where I think part
27:14of what differentiated the analytical
27:17team that I worked on there was that we
27:19saw ourselves not only as analysts but
27:22also as coach and consultant to the
27:26folks who we were working with to
27:27provide data and drive insight and
27:30action so I think one of the challenges
27:32for anyone who works with data is you
27:34show some data and then somebody asks
27:36for more data and then they ask for more
27:38data and more data and more data and you
27:40can get this death by data I think
27:44sometimes this comes about because of a
27:45false hope on the requesters side that
27:48more data is going to answer the
27:50question for them or tell them how to
27:51act when really at the end of the day
27:53it's a person always who's making that
27:56decision but one question that I found
27:58myself often asking that would be useful
28:01and you can think about what the right
28:02framing is depending on your role and uh
28:05the person with whom you're speaking but
28:08help me understand how this new data is
28:11going to help you make a different
28:13decision or a better decision and you
28:16can frame that in the with the tone of
28:18you know help me understand so that I
28:19can better meet your needs but
28:21oftentimes as you can get your audience
28:23or get the people requesting data to
28:25articulate that and really think through
28:26their logic for asking for more data
28:29sometimes they can talk themselves out
28:31of the need for more data uh which can
28:34be useful way to Cull uh this death by
28:38data that we've talked about there
28:40actually was a sort of related example
28:42at a workshop recently where we were
28:44discussing and there was actually this
28:46aha moment of understanding that uh you
28:49could see happen during the discussion
28:51and it was where the team recognized
28:53that they they've historically seen
28:55themselves and SE seen their role as
28:58being primarily to inform and the
29:00discussion was about how they can move
29:02past that U both themselves but then
29:04also in the perception in the
29:05organization to not just being there to
29:08inform but there to actually influence
29:12and drive action and I think there's a
29:14really important pivot that happens when
29:16as analysts we think of our role in that
29:18way not simply to inform but to
29:20influence and drive action and help
29:23people both understand things better but
29:26understand them in the context of now
29:27how they can make changes or how that
29:30reinforces what's been happening from a
29:32business standpoint and I think for me
29:34this is a place where analytics often
29:36stop short where it's easy to show data
29:39and outline some findings but the
29:41challenge or the risk in doing so is our
29:43audiences are faced by a ton of data so
29:46when we give them more data it's a very
29:48easy reaction uh for the audience to say
29:50oh that's interesting and then move on
29:52to the next thing whereas if we take it
29:54to that next step and say not only here
29:56audien is the data but here is how you
29:59should act based on this data or here is
30:01a menu of potential actions you could
30:03take based on this data it gives our
30:05audience something to react to and that
30:08often starts a conversation even if they
30:10disagree uh and that's a conversation
30:13that often gets missed if we stop by
30:16simply showing data so I think always as
30:19an analyst thinking it through to not
30:21only what do I want my audience to know
30:23but what do I want them to do with that
30:25and how do I make that clear and so it's
30:28combination of being coach consultant
30:30helping people understand when and why
30:32they need more data but then also
30:34helping them think through the actions
30:37and really think of our role in
30:39providing data is not simply to inform
30:41it's to help someone understand
30:43something better so that they can make a
30:45smarter decision or take a better
30:49action Alice from Melbourne says I want
30:52to ask a very simple question if someone
30:55like me needs 101 material on data Vis
30:57what sources do you recommend are there
30:59any books that are considered data viz
31:01gospel so I was scanning my bookshelf in
31:04uh figuring out how to answer this one
31:07and I think one that I'll recommend is
31:09Donna Wong's The Wall Street Journal
31:11guide to information Graphics this is uh
31:13it's short it's accessible very easy to
31:16navigate and covers the basics so
31:18different types of charts uh but then
31:20also primers on how to do the math
31:23correctly things like how do you
31:25calculate percent change uh so this
31:27would be a good one for anyone just
31:29getting started and wanting some basic
31:32guidance uh another book one of my early
31:34influences was Steven fuse show me the
31:37numbers so this is more textbook like uh
31:40covers graphs and tables but then also
31:42goes beyond that and gets individual
31:44perception and introduces the idea of
31:47helping your audience know where you
31:49want them to look in what you show when
31:53it comes to data viz gospel certainly
31:55Edward tufty the visual display of Quant
31:58ative information would fall into this
31:59category this is a beautiful and
32:02inspiring book uh perhaps a little less
32:05pragmatic than some of the others that
32:07I've mentioned uh I will also mention my
32:10book storytelling with data which
32:12combines data visualization best
32:13practices with storytelling for getting
32:16information across effectively in a
32:18business
32:19setting all right so I received many
32:22more questions than I can cover in a
32:25single session so if you asked one that
32:28wasn't answered here please listen for
32:29it in future podcasts if you have
32:31questions you can email those to ask
32:34Cole at storytelling with data.com for
32:36potential future inclusion I want to end
32:39with a quick tip here uh this was one
32:41that came up at a workshop a couple of
32:43weeks ago and when it was stated I said
32:46oh my goodness I can't believe I never
32:47knew that which is if you're working in
32:50PowerPoint and you save your file as PPS
32:53as in Sam it will open the file directly
32:57into presenter ENT mode for your
32:59audience to whom you're sending it so
33:01try that out if that's something you
33:04desire and with that I'll say a very big
33:06thanks for tuning in you can get much
33:08more related content by following the
33:10blog at storytelling withd dat.com uh
33:13you'll get regular tips and tricks and
33:16articles and such from me by following
33:18on Twitter and Instagram @ storywith dat
33:21you can like us on Facebook and do all
33:23that great stuff so thanks very much for
33:25joining and I hope you'll tune in again
33:27soon
33:29[Music]

About this transcript

This page contains the full transcript of storytelling with data: #1 the art of feedback by storytelling with data, generated from the public captions YouTube serves with the video. The transcript has 5,825 words across 834 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.