YT YouTube2Text

YouTube2Text → Transcripts → How to turn data into stories

How to turn data into stories — Transcript

by storytelling with data · 7,606 words · 1,126 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Do you ever need to turn to multiple
0:04people or different systems to get data?
0:08Hi, I'm Cole from Storytelling with
0:10Data. And in today's mini workshop, that
0:13is the challenge that we're going to
0:15address as well as how to overcome it. I
0:19should mention that we are broadcasting
0:21for the very first time from our new
0:25production studio at Storytelling with
0:27Data headquarters in Milwaukee,
0:29Wisconsin. And it's been a bit of time
0:32since we've done an open to everyone
0:35live event like this. So, we weren't
0:38really sure what to expect. The team and
0:40I were thrilled to see all of the
0:44excitement building around this. And I
0:47will say as we've been waiting to get
0:48started, I've been having so much fun
0:50monitoring the comments and seeing all
0:53of the places around the world that we
0:56have people tuning in from. Today we had
0:58more than 10,000 people register for
1:02this event. And I just love the fact
1:05that so many people want to learn with
1:08us because we very much enjoy learning
1:11with you. I thought to do something fun
1:13that we would start out by recognizing
1:15the 10,000th registration and that was
1:20Johnny Weathersby who's joining us today
1:23from I assume sunny San Diego,
1:26California. Hi Johnny. Uh we'll be
1:28following up with you to have you select
1:30some fun storytelling with data swag
1:33from our shop. I'm also going to stick
1:35in the mail for you signed copies of all
1:38three of our books. Speaking of books, I
1:42know folks are excited for the many more
1:46that we've promised to give away.
1:48Hundred in fact. Stay tuned. We'll uh
1:51share the winners of those a little bit
1:53later in the hour. For those tuning in
1:57live, you have the opportunity to
1:59participate throughout the session
2:01today. You'll do that by sharing your
2:03ideas in the chat window. I encourage
2:06you to put any questions you have there
2:08as well. The team is monitoring that and
2:12we have some dedicated time for
2:13questions a little bit later in our hour
2:15as well. With that, let's jump in.
2:21Imagine that I am the HR business
2:24partner for my company's sales
2:26organization. I've been invited to an
2:30upcoming leadership offsite and asked to
2:33give an update on the sales manager
2:36population from a people perspective. I
2:39decide to focus on the aspects that are
2:42relevant to headcount hiring,
2:44promotions, transfers and attrition. Now
2:48I am not the data person in this
2:51particular scenario. So I turn to the
2:54people who are and it turns out it's
2:59different people because the data lives
3:01in different systems. And so I know that
3:05I'm eventually going to need to pull
3:07this all together into something that
3:09looks cohesive. So to try to make that a
3:12little bit easier, I provide a template
3:14to my colleagues. This dictates things
3:17like fonts and colors. I even thought
3:21the size of the graph would be
3:23important, so I put a placeholder in for
3:26that. Take a moment and look at what my
3:29colleagues shared.
3:32headcount,
3:34hires and promotions,
3:37internal transfers into and out of sales
3:40manager positions, and finally sales
3:44manager exits from the company.
3:47I think this is a good juncture to
3:50invite some interaction as we think
3:52about how we could pull these graphs
3:55together. I'm going to prime those
3:58tuning in live to get your chat ready.
4:00And I want to know what you think about
4:04this. What is your reaction to this
4:07slide? If you had to describe how it
4:10makes you feel in a single word, what
4:14word would you use? Let me know via
4:16chat.
4:18This is a common approach. By the way,
4:20we have four graphs. So, let's simply
4:23put them together on a single slide.
4:26Sometimes we're even constrained to this
4:30when somebody tells us, "Put it all on
4:33one slide. Give me a comprehensive
4:35view." This unfortunately often leads to
4:39some suboptimal design decisions and a
4:43communication that might be dense with
4:45data but doesn't actually satisfy
4:48anyone. Now, I can see out of the corner
4:51of my eye the chat window is going
4:52crazy, busy, confusing, lots of clutter,
4:56boring, overwhelmed.
4:59These are not the sorts of reactions
5:03that we want to be prompting in our
5:06audience. So, I'm going to suggest that
5:09there is a better way to communicate
5:11this data. Desperate data is a common
5:15challenge, but just because we start
5:18there does not mean we are destined to
5:21end with disperate data. Today, I'd like
5:24to invite you to accompany me on a
5:27journey. Going to take that desperate
5:29data and start by turning it into some
5:32good graphs. We will take a couple
5:35straightforward steps to transition
5:38those good graphs into something great.
5:41and then we are going to take a great
5:44leap forward and turn those great graphs
5:47into a stellar story. Let's get started
5:52with good graphs. Two tips here. First
5:56is to cut the clutter. Second is to make
6:00the details consistent. Let's start off
6:04with a conversation on clutter. I'll
6:07bring back one of those graphs I flashed
6:09in front of you a moment ago and ask you
6:12to help me decide what clutter we can
6:16eliminate from this graph. And now I
6:18think of clutter simply as elements that
6:21are present in our visual communications
6:24that don't need to be. When you imagine
6:28that every single element we put on a
6:31graph or a slide, it creates density. It
6:36brings a burden cognitively to our
6:38audience. Want to make sure all of those
6:41elements earn their place. Uh again, I
6:45see chat going crazy with things that
6:47people want to eliminate from this
6:50graph. Grid lines, labels, the vertical
6:54axis,
6:56uh shorten the names on the xaxis. I
6:59think that is a fantastic idea. Uh the
7:03chart border. Yeah, a lot of people
7:05commenting on grid lines. I also see a
7:08question posed from Martin. What
7:10matters?
7:12We'll get there. Bear with us. But
7:14first, let's do some decluttering.
7:18Let's start with the easy ones. I'll
7:20take away the graph border and the grid
7:23lines. It's always amazing to me how
7:25much those two steps alone do in terms
7:28of making my data stand out more. Next,
7:33I'm going to clean up my Xaxis labels.
7:37And diagonal labels, you know, they
7:38maybe aren't the end of the world, but
7:40they also aren't great. They look
7:42sloppy. They create this jagged line at
7:45the bottom of our graph. But worse than
7:47that, studies have shown diagonal text
7:50is about 50% slower to read than
7:53horizontal text. So, as I see a lot of
7:56people bringing up in the comments, we
7:59can simply shorten those to the
8:01abbreviation. also creates some nice
8:03clean structure along the bottom of our
8:06graph. Next, and this is maybe more of a
8:09pet peeve than anything, but it just
8:11gets under my skin when the white space
8:14between the bars is bigger than the bars
8:17themselves. Can feel sort of visually
8:20jarring. So, I'm going to thicken up
8:22those bars. And actually, one
8:24opportunity this affords me is if I want
8:27to keep those data labels, bear with me
8:30for a moment because I will get rid of
8:31them. Uh, but I can pull them into the
8:34ends of the bars. This is a really cool
8:37trick because this reduces the perceived
8:40density of what I'm showing without
8:43actually reducing any of the
8:45information. But as many people are
8:49commenting, we don't need both our yaxis
8:53and every single data point labeled. So
8:55I can choose one or the other of those.
8:59And typically when you're making that
9:00decision, what you want to think about
9:02is how important are the specific
9:04numerical values. If they're critical,
9:07then you can leave them there and omit
9:10the axis. If on the other hand, you'd
9:12rather people focus on the general shape
9:14of the data or comparisons across
9:16different data sets, then often times
9:19you don't want to clutter the graph with
9:20that and can instead preserve the axis.
9:23That's what I'm going to do in this
9:25case. As some additional cleanup, I'm
9:29going to orient my titles at upper
9:32leftmost. This creates some nice visual
9:35framing for my graph. Also, it means
9:37that people hit how to read the data
9:40before they get to the graph, which is a
9:43nice thing. When it comes to reading
9:45this, I can do some things to make that
9:47slightly easier. Make my yaxis title a
9:50little shorter and piffier. I'm also
9:52going to add some bolding to my title so
9:56it's a little more scannable. I'm going
9:58to make just one more change at this
10:00juncture, which is to lighten things up
10:03by turning this data over time into a
10:07line graph. Let's take a look at where
10:10we started. Now, each of these changes
10:14on its own totally minor, but when you
10:17layer them together, these individual
10:20small changes that are easy to make
10:22happen have great impact.
10:26Let's go back to that original view and
10:30ask for you to help me spot some
10:33inconsistencies.
10:35What inconsistencies do you see? Where
10:38are things different? Where they could
10:41be the same? Those tuning in live can
10:44let me know via the chat window. This is
10:47another aspect that can make our graphs
10:50unnecessarily harder to interpret. It
10:53also displays a lack of attention to
10:56detail. Taking a few minutes to make
10:58anything that can be consistent across
11:01similar views the same makes things
11:04easier for our audience. see what you're
11:07highlighting in chat. The xaxis
11:11colors I see coming up a lot of time.
11:14Capitalization.
11:16Yes, it's different uh in a number of
11:18places there where it doesn't need to
11:21be. Uh
11:24yeah, lots of great ideas coming up
11:27here. Font size and case, uh yaxis name,
11:31the legend in different spots. Yes, all
11:34of these things. And again, each
11:36individual one minor, but together they
11:39create a less than spectacular feeling
11:42when it comes to how our audience
11:44interprets our work, which is not what
11:48we want. Let's just look at a few of
11:49these. I'll highlight them sequentially.
11:52There a lot of inconsistencies. So,
11:54every single graph title is approached
11:57differently. Even though they all use my
11:59predefined font, they're different
12:01sizes. There's a different case
12:03structure approach. even the words that
12:06people chose across the different titles
12:08uh varies. We can do a lot to simplify
12:11that because when things could be the
12:14same but aren't, it causes our audience
12:17to question why that is, which is not
12:20where we want them spending their brain
12:22power. We want them spending their brain
12:24power to understand what we want them to
12:26see and understand and what to do with
12:29that. So, we can bring some consistency
12:31to our graph titles. While we do that,
12:33let's bring some consistency to our
12:35yaxis titles as well as many people have
12:39suggested. All of these are some
12:42component of the sales manager
12:44population. So we can make those titles
12:47similar or the same across the various
12:50graphs.
12:51Months of the year though they run
12:53consistently from January to December
12:56across the graphs, they're approached in
12:58a slightly different way across both of
13:00them. So you can just bring consistency
13:03there as well as in the legend
13:06placement. So where the legend is
13:08present, it is in a different spot on
13:11every single graph. When it comes to
13:13legend placement, I'm a fan of labeling
13:16data directly when you can. When you
13:18need a legend though, think about also
13:20orienting that at the upper left uh
13:23potentially under the graph title again
13:25so your audience hits how to interpret
13:27the data before they get to the
13:30specifics.
13:32Going back to the original Oh, colors.
13:36Uh lots of colors are used here. They
13:38were the ones I prescribed. Uh but
13:40they're not used very thoughtfully.
13:41We'll address that. So going back to the
13:45original and taking these two steps
13:48together, we can move from cluttered,
13:51inconsistent,
13:53visually disperate data to clean and
13:57consistent good graphs.
14:01However, we do not want to stop here.
14:05When we only take away, people can feel
14:09like we've stripped things out and not
14:12added back value in its place. And it
14:15turns out there are two simple steps we
14:19can take to those good graphs to make
14:21them great. We can focus attention
14:25sparingly and use words wisely. Let's
14:29take a look at how we can achieve that.
14:33Let's look at one of these stripped
14:35down, decluttered versions of a graph
14:38and talk about how we could focus
14:41attention on one of these lines. Now,
14:46I've started out by intentionally
14:48pushing everything to the background,
14:49making it gray. This gives us the
14:52ability to achieve visual contrast. And
14:56sparing visual contrast is going to be
14:59how we signal to our audience where we
15:02want them to look. See what's coming in
15:06via chat.
15:08I see a number of people talking about
15:11talking about color as a way to
15:14differentiate. We absolutely can. Color
15:17used sparingly is one of our most
15:20powerful tools for directing attention.
15:23But there are other things we can do to
15:24create contrast as well. Let's assume,
15:27for example, that we want to direct our
15:29audience's attention specifically to the
15:31higher line. This is the one that starts
15:34off the highest. It has clear peaks and
15:37valleys over the course of time and it
15:39also ends the highest. What besides
15:42color could we do to that line? See,
15:45Johan says intensity
15:47uh line types comes up. We could think
15:50about making where we want people to
15:52look a dashed or dotted line. See a
15:55couple of other votes for that line
15:58thickness. Ah, from Johnny Weathersby,
16:00our lucky 10,000th registration. Thanks,
16:03Johnny. See a number of people
16:06suggesting that we make that line red.
16:10Ah, Neo suggests words or phrases. Those
16:14are definitely going to come into play.
16:19There are a lot of ways that we can show
16:23our audience where we want them to look
16:25through sparing contrast. Let's take a
16:28look at some of those. Color is probably
16:31the most obvious one. One thing I will
16:35mention with color is it's unique as a
16:38design aspect in its ability to impart
16:43tone or feeling on the things that we
16:45put in color. So notice here that blue
16:48feels positive, friendly, nice. Whereas
16:51if I simply take that same line and I
16:54make it red, now it feels like danger or
16:58aggressive, something bad might be
17:01happening. So, we want to consider how
17:03we can use color and the tone that it
17:06can impart to reinforce what we want to
17:10get across and to make consistent things
17:14consistent visually through the similar
17:16use of color. We'll see that play out in
17:20a moment. Terms of other ways to direct
17:22attention to that hire's line. As a
17:24number of people noted, we can make it
17:26thick, make the other lines thinner, or
17:29a combination of those things. Intensity
17:33is something else we can play with. We
17:34can make the higher line darker than all
17:37the rest. Now, before I flip to the next
17:40one, which is position,
17:42we can't move the line around. If we
17:45have other graph types, sometimes that's
17:47possible to resort how we're showing the
17:49data. But here, the line is where it is
17:51because of the data it's plotting. But
17:53we can ensure that it doesn't cross
17:56behind other data series. So if you
17:58direct your attention to October and the
18:00space around it, we see another one of
18:01those lines crossing in front of it. So
18:03we can just bring the higher line
18:05visually forward so that we don't have
18:07that issue. Dotted lines stand out very
18:11much when other things are not dotted.
18:14I'm a big fan of dotted lines to express
18:16uncertainty. It's a goal, a target, uh
18:19an estimate of some point.
18:22We could remove all of the other data.
18:25I'm sure that came up somewhere in chat,
18:27but I didn't see it specifically. This
18:29is always something we want to ask
18:30ourselves. By the way, do we need all of
18:34the data that we're showing? When you
18:38consider eliminating data, however, be
18:40thoughtful of what context that you lose
18:43when you do so and make sure that that's
18:45an appropriate trade-off. On the flip
18:47side of this, we could show just the
18:50other data series and then have the
18:53hires line appear. And that simple
18:56animation of it not being present and
18:59then becoming so garers attention. And a
19:03live presentation that can be a really
19:04useful thing to do. I saw comments for
19:08data markers and data labels. Yes, if we
19:11put them everywhere, we might end up
19:13with a cluttered mess. But we can
19:15actually incite our audience to make
19:18specific comparisons when we are sparing
19:21and considerate about which data labels
19:25we include. For example, if I highlight
19:28just these peaks, we can start to say
19:31words about this graph. We might say
19:34something like hires tend to happen most
19:38in the first month of each quarter.
19:41Those words are important. If that's
19:45what we want our audience to know, we
19:48should put them on the page or on the
19:52graph. There was actually one prominent
19:55study recently that showed when you
19:57title your graph like this with the
19:59primary takeaway, people are more likely
20:02to remember that takeaway. The priming
20:06power of words is really, really useful.
20:11And when we pair that with the sparing
20:14visual contrast, then we've done some
20:17really nice things for our audience,
20:19which is we've made it clear where to
20:22look. And through our words, we've made
20:25it clear what to see.
20:28So you can imagine how we might do this
20:31for each of the other graphs. Rather
20:34than do that, however, I want to show
20:36you how we can take things a big step
20:40forward through story. Uh, but before we
20:44get there, I want to share a few
20:46additional ways that everyone can learn
20:49with storytelling with data.
20:52We have just launched our 2024 public
20:57workshop schedule. Today you're seeing a
21:00sampling of strategies that enable us to
21:03communicate effectively with data. You
21:05can learn even more in these sessions
21:09and we have a variety of them to meet
21:11your individual needs. Uh from the short
21:13punchy storytelling with slides focused
21:16on planning presentations and designing
21:19stellar slides. We have our storytelling
21:21with data classic workshop that dives
21:23deeper into content similar to what
21:25we're covering here. uh making effective
21:28graphs, weaving them into action,
21:30inspiring stories, uh but goes quite a
21:33bit more in depth given the longer time.
21:36We also have a one-day master class that
21:38combines all of that great learning plus
21:42more on how you can deliver a stellar
21:45presentation. Those are in person and we
21:48have sessions planned in April in London
21:51and in September in Seattle. And then
21:54finally, for those who want to learn
21:56even more in a longer uh format, we have
22:01our 8week online course and there are
22:04cohorts of that starting in January and
22:07again in the fall. I will mention for
22:09those tuning in live or watching this
22:12video later, you can use the code good
22:15to great. Uh, that's g o d t og gre a t
22:21at registration for any of these
22:23sessions for 10% off. You'll find all of
22:26the details at
22:27storytellingwithdata.com/workshops.
22:30I'll also mention we are going to be
22:33rolling out an official scholarship
22:35program for all of our 2024 sessions. So
22:38stay tuned for that. In the meantime,
22:40folks can register with that good to
22:42great code for 10% off. Also want to
22:46highlight our custom sessions. If you
22:49want to organize or suggest learning for
22:52your team or organization, we offer
22:54private and custom versions of various
22:56sessions ranging from shorter inspiring
22:59keynote presentations and skill-building
23:01webinars to longer form workshops where
23:04we collect examples from your team ahead
23:06of time and use those to illustrate and
23:08practice the lessons covered. More info
23:11about these offerings can be found at
23:13storytellingwithdata.com/custom-workshops.
23:19I'll also mention that for organizations
23:20who want to learn with us but are facing
23:23constraints. We also have a special
23:26program called reach. This is
23:28applicationbased. So you can apply to
23:30bring lowercost sessions to your team.
23:33You can find information on that at
23:35storytellingwithdata.com/reach.
23:38All right, I'm almost done with the
23:40advertisement part of our session, I
23:43promise. Just want to draw your
23:45attention to all of the other resources
23:47that we make available. And like this
23:50mini workshop today, a great deal of the
23:53content that we produce is free and open
23:57to everyone from videos, our blog
24:00articles, podcast episodes. We also
24:04offer ways to practice and exchange
24:05feedback in our online storytelling with
24:07data community. That's also where we
24:10start to get in some ways you can
24:11support us as well through premium
24:13subscription there or as I mentioned by
24:15booking a workshop for yourself or your
24:18organization
24:20uh reading our books. Actually, on that
24:22front, I'll mention that we have a
24:24couple of fun projects underway that are
24:27taking shape in this space, including
24:29something that those out there who both
24:32work with data and have children in
24:35their lives will appreciate. Stay tuned
24:38for more on that front. I should also
24:41mention when it comes to books, we are
24:45giving away a hundred copies of my
24:48newest book, Storytelling with You,
24:51Plan, Create, and Deliver a Stellar
24:55Presentation.
24:56And we'll go to the slide where you can
24:59see who those lucky 100 recipients are.
25:05I'll pause here for a moment. Uh
25:08congratulations to everyone who will
25:10have a book coming to them after the
25:12session here today. We'll say for
25:15everyone else, you can get storytelling
25:17with you or pick up any of our books at
25:20your favorite retailer.
25:25Before I get back to our content, I want
25:27to just give a quick reminder that we
25:29have time set aside today for viewer
25:32questions. So for those tuning in live,
25:34please share your questions on any topic
25:37related to our content today or really
25:40anything related to making effective
25:43graphs and giving powerful presentations
25:47would be welcome. You can share those in
25:49the comment or chat window.
25:53First, let's finish this up with a
25:56stellar story. And to make stellar
25:59stories, we want to do two things.
26:01First, weave multiple graphs together
26:04and secondly drive people to do
26:08something specific with the data that we
26:11share, enticing them to act. Let's take
26:14a look at what this can look like.
26:18I'm Cole and I'm here today with a sales
26:20manager update and I want to encourage
26:23us to rethink how we hire sales managers
26:27in the organization.
26:29Just to set the stage, I'm going to be
26:31looking at our sales manager population
26:34and the aspects that contribute to
26:36headcount over time. Look at this for
26:39the last calendar year aggregated
26:41together. So just directing your
26:43attention down to the bottom, that
26:45xaxis, I'm going to start with the
26:47beginning of year headcount. Then I will
26:50add in the additions to headcount,
26:53hires, promotions, transfers in to sales
26:56manager positions. And then we'll take
26:59away the deductions to headcount
27:02transfers out of sales manager positions
27:04and exits from that population. This is
27:07going to basically be like a visual math
27:09problem where eventually it will yield
27:12the endofear headcount.
27:15We started the year with 317
27:19sales managers. The biggest addition by
27:22far was through hiring. uh hiring
27:25actually accounted for 2thirds of the
27:28growth to this population over the
27:29course of the past year. We did also
27:32have promotions and transfers in. These
27:35count accounted together for that
27:36remaining third of increase to our sales
27:40manager population. We also had some
27:42deductions transfers out and I'll just
27:45bring attention to the fact that we had
27:47more transfers out than we had
27:50promotions and transfers in combined. We
27:54also had a great deal of exits over the
27:57course of the past year. You'll note
27:59that we had more exits than we hired.
28:02So, taking all of this together means
28:05we're actually slightly down
28:08year-over-year on headcount from where
28:10we began. While at the same time, the
28:14overall sales organization has grown,
28:17meaning we have an increased need for
28:19sales managers. Now, you might simply
28:22think, well, that means we should hire
28:23more. But I'm going to suggest a
28:26different approach. First, I want to
28:28share some additional detail on how
28:30these metrics play out over the course
28:33of the year because I think this can
28:36help guide our forward-looking strategy.
28:39So, let's focus focus first on the
28:41additions to headcount. Going to go to a
28:45different structure here. We're still
28:46looking at the number of managers on the
28:48y ais, but now we have months over the
28:51course of the year from January to
28:52December on our x-axis.
28:5567% were hires. Uh in other words, every
29:00two out of every three new sales
29:03managers came into the organization from
29:06the outside this past year. I'll just
29:10highlight the fact that we have the
29:12greatest number of hires starting the
29:14first month of each quarter. That's
29:16largely due how we set our targets and
29:19sales incentives.
29:22You'll note that there are relatively
29:24fewer hires when it comes to the first
29:26month of the quarter in October and that
29:28is due to our annual promotion cycle
29:30that takes place then. So let's jump
29:32next to looking at that trend. We had 47
29:35promoted over the course of the last
29:37year to sales manager. 38 of those
29:41happened during our single annual
29:43promotion cycle. Now, I'll just mention
29:45when it comes to promotions, couple
29:48great things about them is we already
29:51know that folks getting promoted into
29:54sales managers are a culture fit with
29:56the organization in general and with the
29:59sales organization in particular. But
30:01perhaps even more important than that,
30:04those promoted from within tend to stay
30:07in role and at the organization longer
30:10than external hires. On the flip side of
30:14promotions and some of the challenges
30:16there with the single annual cycle, we
30:19hear complaints every year from the
30:21existing leadership team. It's just all
30:24consuming in that time period. And also
30:28anyone who's just shy of being ready to
30:31promote has to wait an entire additional
30:35year before they become eligible again.
30:38And we actually lose many people during
30:40that time period.
30:42Before we move to transfers out and
30:45exits, take a look at transfers in. So
30:48similar to hires, we tend to see more of
30:50those in the first month of each
30:52quarter. Similar to promotions, we get
30:56the benefit of culture fit and again
30:58longer time in role and at the company
31:01than we see with external hires. Now,
31:04you might be able to anticipate where
31:07I'm going to be going with this, which
31:09is if we take some of that energy that
31:11we've historically spent on hiring and
31:14use it instead to increase our
31:16promotions and transfers, I think it
31:19could be a win across several fronts.
31:22Before we talk more about that, let's
31:24take a look at the deductions from a
31:26headcount. Those come in the form of
31:29transfers out of the organization and
31:32exits. So again, let's look at those
31:34over the course of time. So I'm back to
31:36that view with uh sales managers on our
31:39y-axis and months of the year on our
31:41x-axis. We had 111 transfers out. I'm
31:45going to go ahead and layer on exits as
31:48well because these follow a similar
31:50trend over the course of the year.
31:53January is when many of the hires from
31:57the prior year become eligible to
31:59transfer and as we can see many of them
32:02do. Also digging into exits in January
32:06and February we can see that many of
32:09these first pursued internal transfer
32:12but were unsuccessful. And again, many
32:15of these were more recent hires, meaning
32:18maybe they see it as a foot in the door,
32:20but aren't thinking about sticking
32:22around. We also see a spike in both
32:25transfers and exits in November. Anyone
32:29know why that might be?
32:34That's right. These are people who maybe
32:37thought they were ready for a promotion
32:40but didn't get it. And so we see exits
32:44coming through there. Just back to the
32:47big picture and taking all of this into
32:51account, it may be time to consider
32:53altering our people strategy when it
32:56comes to sales managers. If you consider
32:58the effort it takes to bring in this
33:01many hires, we think that shifting some
33:04of that energy towards increasing
33:07promotions and transfers into the
33:10organization will help us reduce those
33:13red bars of sales managers transferring
33:15out and leaving the organization.
33:19Just to take that and put it into words
33:21on a slide, I'm recommending that we
33:24increase our efforts to hire from within
33:26the organization. That can take a number
33:29of different forms. We could fasttrack
33:33internal transfers into sales manager
33:35positions, making that a fast and easy
33:38process. I highly recommend that we add
33:41a second promotion cycle in April, both
33:45to spread out the burden on the current
33:47team as well as give people line of
33:49sight to that next potential promotion.
33:53We could also consider introducing an
33:55approval process for offcycle
33:57promotions. And I'm sure that you have
34:00ideas as well. Let's discuss and
34:03determine where we go from here.
34:08That was a great experience and in fact
34:13after going through all of that it is
34:16possible to get the key pieces into a
34:19single slide summary. I wouldn't present
34:23this but it can be used as a follow-up
34:26to remind people what was covered or for
34:28those who missed the update. You may
34:31find it difficult to recognize now that
34:35this is where we started.
34:38We improved things quite a bit. As I
34:42mentioned, just because we began with
34:45disperate data did not mean we were
34:47destined to end there. We started by
34:49turning it into good graphs by cutting
34:52the clutter and making the details
34:55consistent. We made those good graphs
34:58great by focusing attention sparingly
35:01and using words wisely. And then we took
35:05a giant leap forward into a stellar
35:09story, weaving multiple graphs together
35:12and driving people to act. So the next
35:15time you find yourself facing desperate
35:19data, reflect on the lessons we've
35:22covered here. Use them to make great
35:24graphs and move beyond just showing
35:28data. Make those great graphs a pivotal
35:31point in an overarching story.
35:37It's time next to turn our attention to
35:40questions. If you haven't already, I
35:42invite those watching live to please ask
35:45your questions in the chat window. Randy
35:48has been here behind the scenes so far
35:51making sure things run smoothly. Randy,
35:54do you have any viewer questions for me?
35:58Sure thing. So, uh, first we had a
36:00question way back that I made note of
36:02when you were talking about the
36:03different books and that question was
36:05from Panda who asked, "What's the
36:07difference among the three different
36:09books?"
36:11Fantastic question. So, I'll start with
36:12the original, that's the white book,
36:15Storytelling with Data. If you're
36:17working with data on a dayto-day basis
36:20and need guidance for how you can turn
36:22that into effective graphs in terms of
36:25what type of graph to use, more examples
36:29on how you might declutter and focus and
36:32weave things together into a story. This
36:35is a great place to start. Uh basically
36:38goes deeper into a lot of things that we
36:40talked about today and with many more
36:43examples. You'll find even more examples
36:47in Let's Practice. So, this book is
36:49structured chapter and lessonwise along
36:52the same lessons as the original book,
36:55but it is entirely exercise-based. And
36:58so, within each chapter, there are three
37:01sets of exercises. First, there's
37:03practice with Cole, where I put forth a
37:06scenario that you're meant to think
37:07through and maybe approach on your own,
37:09but then I also show you how I would
37:11approach it. It's a way of getting
37:13insight into many more examples and
37:17corner cases and all the issues that
37:19come up when we're grappling with
37:20graphing and communicating data. There's
37:23another section of exercises called
37:25practice on your own which are canned
37:29sort of examples but without any
37:31prescribed solutions. These are great
37:32for university instructors teaching from
37:35the books. uh we have I think over 500
37:38or 600 now identified of instructors
37:41around the world teaching for our book
37:42from our books. So you can use those for
37:44additional homework or group projects.
37:46Also great for the individual who just
37:48wants to learn more or for a manager of
37:50a team who might want to encourage that.
37:52And then the final exercise section
37:54within each is practice at work where it
37:57takes the concepts and really breaks
38:00them down into guidance about how you
38:02might take something you're facing in
38:03your job and apply the lessons. This is
38:06great for somebody who wants a handson
38:10way to learn. And then I will say the
38:13one I'm most excited about at this
38:14moment in time is the newest
38:17storytelling with you which goes beyond
38:20the other two books and really gets into
38:23the important role that the individual
38:26plays when communicating whether it's
38:28data or anything. Because you can make a
38:32great graph, but if you can't talk about
38:35that graph or that data in a way that
38:38engages and gets people to want to
38:40listen and act, the beautiful graph or
38:44slide is going to fail. So I often get
38:46asked, what should I do next after I
38:50read the first book or take a workshop?
38:52And you know, I want to I want to make
38:53even better graphs. But I would say
38:56don't worry about better graphs. Graph,
38:58good graphs, great graphs, that's good
39:00enough. The next way to really advance
39:03yourself is to invest in yourself and
39:07how you present yourself, how you
39:09present your data, how you talk, how you
39:12engage. And so the new book really walks
39:14through that. Um, as well as the
39:17practical bits of creating and planning.
39:20Uh so it takes you through getting clear
39:22on your message, understanding your
39:24audience, planning out your content in a
39:26low tech way, then goes through the
39:28technical aspects of bringing that low
39:30tech planning into your tools, going
39:33through things like setting up a
39:34template in PowerPoint to make things
39:36easy and consistent. There's an entire
39:39chapter on graphs, uh also chapters on
39:42words and images as well. And then the
39:45final section really dives into
39:47developing yourself. So, I would say if
39:49you're debating which do I get,
39:52start with this one.
39:55All right. And there was a question of
39:57which format are those books in? And
39:59actually, they're all in electronic
40:01format. And then storytelling with data.
40:04The white book and the yellow book are
40:06both available on Audible, read by the
40:10author, which is which is always
40:12exciting. All right, we had another
40:13question. Uh
40:16uh many people are asking what tools do
40:19we use to make our graphs and criti
40:23also ask how do you animate your charts
40:26in those tools?
40:28Great questions. Everything that we've
40:31seen today was done directly in
40:33PowerPoint. And I will say the majority
40:36of what the team and I do is PowerPoint
40:39or a combination of Excel and
40:41PowerPoint. mainly because these tools
40:43are pervasive. Love the fact that anyone
40:45can pick them up and make a graph,
40:48right? There's no barrier to entry.
40:50Challenge is just that nobody really
40:52teaches us how to do this. So, the kinds
40:55of lessons that we focus on across all
40:58of our work are those that are tool
41:00agnostic that can be achieved in any
41:03tool. So, when it comes to tools, I'm a
41:06fan of picking one or a couple and
41:09getting to know them well so that they
41:11don't become limiting when it comes to
41:14employing some of the things that we've
41:15talked about today when it comes to the
41:18individual questions of how did you do
41:20that in PowerPoint. So, for what we saw
41:23here, it's a lot of the same graph on
41:26different slides just formatted
41:27differently, which creates that animated
41:30feel as I flip through them. And a great
41:33resource for you to turn to on that is
41:35the Storytelling with Data YouTube
41:37channel uh because we have a ton of
41:40tutorials and more coming uh and a lot
41:42of shorts as well that will show you
41:46what menu uh settings to go through when
41:48it comes to some of those formatting
41:50changes and how we actually go through
41:52and animate in these sorts of settings.
41:54So definitely recommend checking out
41:55resources there. We will also follow up
41:58with everyone who registered for the
42:00session today and make sure that we
42:01include all of the resources that we
42:03talk about here.
42:05All right. In a related question, Diana
42:06asks, "What do you do when your audience
42:08requests that you continue to show them
42:11tables for everything?"
42:14The audience who loves tables
42:19is often feeling like their question of
42:23so what isn't answered. And when that is
42:27the case, it feels like getting more
42:28data can be the answer. And so one thing
42:31I would recommend trying though, because
42:33if you simply say, you know what, tables
42:35aren't the right answer, I'm going to
42:36give you a graph instead, people will
42:39not like that because they tend to be
42:41change resistant. So instead of taking
42:43anything away, think about adding where
42:46you can say, audience, I still have your
42:49tables. We can go through those. But
42:51I've done something different today that
42:53I think is going to help us have a
42:55better conversation or see something new
42:58or in a different light. And I've gone
43:01ahead and put some of that data in a
43:03graph. And here's what we can see. And
43:05here's why this is interesting or
43:07important and how it's relevant for you.
43:09And what you'll find is over time as you
43:13start to develop both your own
43:15confidence and your audiences that you
43:18are highlighting the important things
43:20for them. It will wean them off of this
43:23desire for the tables because again
43:26oftenimes people wanting tables it's
43:28thinking that more data is going to
43:30answer the question which means that
43:32their questions aren't getting answered
43:35currently. So if you can get more
43:37context and understand what they need,
43:39how they're making decisions, what
43:41inputs would be useful, that will help
43:44you curate from that tabular data a
43:48story like what we saw today. Also just
43:51look for instances where you are likely
43:53to be successful. Uh so maybe starting a
43:56new project, you might try this instead
43:58of going against the grain of something
44:00that has already been living in a table.
44:02Just a few thoughts.
44:05All right, Brad asks, "Is data
44:06storytelling the same as data
44:08visualization?"
44:10No, it's not. Uh, nomclature is an
44:13interesting thing because words get
44:15thrown around and come to mean different
44:17things over time. For me, data
44:19visualization is simply taking data,
44:22taking numbers, and turning them into
44:24pictures. We can visualize data for many
44:27different purposes. We can do it in a
44:30business setting where we're after
44:32efficacy and the speed of transfer of
44:36information. Uh we can also do data
44:38visualization that is more artistic or
44:42interesting from an aesthetic point of
44:44view. Uh neither of those are wrong or
44:47right. They're just data visualization
44:48for different purposes. Data
44:50storytelling is not just the data.
44:54That's where you are bringing in
44:56components of story. Uh so when we teach
44:59about storytelling in our work, we're
45:02really getting into it. What's the plot?
45:04Uh where is their tension in terms of
45:07what matters to the audience that either
45:09isn't being satisfied in some way or
45:11something that could go wrong? How do we
45:13build that tension over the course of
45:16our data story, reaching a peak of
45:18climax and then having a falling action
45:20and a resolution? So really bringing
45:23structures of story into how we
45:26communicate because when we do that
45:28well, we can use it really powerfully to
45:30engage and get people to stick with us
45:32and get them to care, which is
45:36incredibly powerful. But I will say data
45:39storytelling is one of those buzz
45:41phrases that gets thrown around when
45:42people maybe just mean put some words on
45:44a graph. Uh that's a step towards it,
45:47but there's so much more we can do.
45:50Couple of folks have asked how do you
45:52use branded or familiar colors in a
45:54graph and what do you do when you're
45:56restricted in which colors you can you
45:58can use or as as Liz says um what about
46:01you know when your audience wants to use
46:03red yellow and green and pushes back at
46:05the changes to more accessible colors
46:07and I will add something that Sophia
46:10added which is and what about us
46:12colorblind folks what do you do about
46:14them
46:15yes color as we've seen is an incredibly
46:18powerful tool in our designer toolkit,
46:21particularly when we use it sparingly.
46:23Um, so when there are brand colors that
46:27you can fold into how you're
46:28communicating with data, I recommend
46:31doing that can bring a nice cohesive
46:33look and feel to things. Just recognize
46:36because you have a ton of different
46:37brand colors does not mean you need to
46:39put all of them in your graph. So
46:41picking one or a couple distinct
46:44prominent brand colors and using gray
46:46elsewhere can often work for that. When
46:49it comes to the stoplight question of
46:51the audience who wants the red, yellow,
46:54green, we do get into some colorblind
46:57issues there which might be one argument
46:59that would be useful for your audience
47:02is about 10% of western population
47:05experiences some form of color blindness
47:07which most typically is difficulty in
47:09distinguishing between shades of red and
47:11shades of green. I'd argue also that
47:13mostly when we use those color palettes,
47:15we're not interested in all of it. we're
47:17interested just in what's going well or
47:19just in what isn't going well. So you
47:22could even think of highlighting those
47:23things sequentially
47:26uh instead of all at once. So the
47:28challenge is when everything is
47:30different, nothing stands out. And so
47:34that can be fine if you're using it to
47:35explore the data, but once you've
47:38already done that, you have something
47:39specific you want to communicate and
47:41somewhere specific you want people to
47:43pay attention to, then we want to use
47:45our color more sparingly in order to
47:47drive that. I think we have time for one
47:51final question.
47:54All right, this last question is from
47:56the user handle an SS. Sounds very
48:00mysterious, but the question is a great
48:01one. It says, "What is the best way to
48:03convince leadership that we need to
48:05incorporate storytelling in our comm
48:07communications? Could we say it was uh a
48:10better way to provoke thought or better
48:12position uh position us to make
48:15meaningful decisions and
48:16recommendations?" What are your thoughts
48:18on that?
48:18This is a fantastic question and yes,
48:20all of those things. uh you know it's
48:23hard to point to ROI when it comes to
48:25investing in these skills, but I think
48:28the way that we see them play out is
48:31when it's done and you're finding that
48:33people are having better discussions.
48:36They're making smarter decisions because
48:39they're no longer asking questions about
48:41the data or asking for more data or
48:43trying to understand the graph. they're
48:46able to quickly get to how does this new
48:49information I now have matter for the
48:51business matter for the important
48:53conversations and decisions that we're
48:55having and making and so the more you
48:58can build situations like that and point
49:01to their success. So I would say try out
49:05the things that we've talked today and
49:07that you'll read about in the books and
49:09see on YouTube and elsewhere. Try out
49:12the ones that you think are going to be
49:14the most useful in your work and try
49:17them out in instances where you are
49:20likely to be successful. Uh where the
49:23risks aren't crazy big and people will
49:26be accepting because then you can start
49:28to build momentum. Uh because the best
49:31thing is when people start coming to you
49:33because of your fantastic work. I've
49:35seen what you do when you're
49:37communicating with data. can you teach
49:38my team to do that or can you do that
49:40for me as well? And that's how you get
49:43really great grassroots momentum with
49:46this stuff. So, it won't be successful
49:48every time. Don't get discouraged. Keep
49:51trying. Look for places where things are
49:54successful and build on that.
49:58So, we're out of time. We took the whole
50:00hour and I love it. I love the
50:02excitement. I love being able to see
50:03chat flow through out of the corner of
50:06my eyes. So I just want to say a big
50:09thank you to everybody tuning in today.
50:13Uh this recording will be available. It
50:16will live in YouTube so you'll be able
50:18to rewatch and point colleagues to it.
50:20Also just mention if you enjoyed this
50:23session, please let us know in the
50:25comments. Uh because I think if you do,
50:28we may very well do more of them. And
50:31with that again, thank you for tuning in
50:34today. I wish you great graphs and
50:38stellar presentations.

About this transcript

This page contains the full transcript of How to turn data into stories by storytelling with data, generated from the public captions YouTube serves with the video. The transcript has 7,606 words across 1,126 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 translate the transcript into 100+ languages.

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, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.