Founded in 2010, storytelling with data, states as its mission: trying to “rid the world of ineffective graphs and help people elegantly communicate with data”. You may have visited the website, taken part in a #SWDChallenge, read one of their books or even participated in one of the numerous workshops their team leads around the globe.
I recently had the chance to spend some time with Cole to talk about her latest book, Storytelling with Data: Let’s Practice!, the follow up to 2015’s Storytelling with Data, and to get her perspective on the challenges and rewards of communicating with data.
Charles Saulnier: First off, how are things now that the new book is published?
Cole Nussbaumer Knaflic: Things are good! Things are busy. We’re already working on other exciting resources to help drive effective storytelling out there. Of course the book is out, but there is never a shortage of projects in that space and time.
CS: Now that your team has expanded (with the recent addition of Mike Cisneros and Alex Velez, who join you and Elizabeth Ricks), are you able to stay a bit closer to home and focus on other aspects of your company?
CNK: A little bit more, if I look back at what the last few years have been like. This year early on was still busy, I was for the first time in Australia and New Zealand, and gave workshops in new locations around Europe and the States as well. But yes, the team has grown, we’re now four data storytellers plus a small but mighty crew behind the scenes who help make us all more effective in various ways. The growth is great because it means we’re able to reach more people and help them improve their skills.
CS: Let’s talk a bit about the book. I honestly am not that far into my reading, because there’s just so much to do with it! You mention in its introduction that you wrote it with two audiences in mind: learners and teachers. Since the book came out, did either of these audiences reach out with success stories based on this work?
CNK: From a teaching standpoint, the first book is now used in over a 100 universities around the world. One of the things we’d been thinking of for a while is that we needed to do a textbook out of this, where there are case studies, exercises. And well, if that’s the best way to learn this, then it shouldn’t be a textbook, because no normal person is going to go and buy a textbook, right? So how do we make something that gives people more examples and insights into the work? Because the first book is very cut-and-dry: you found this, therefore you do that. It’s good for teaching, but in the real world, things are stickier, messier, there are corner cases.
The way that Let’s Practice! is structured is to dig into those, and shows the thought-process that goes into it, which is more important than the specifics of a given scenario. The thought behind the three sections is first, Practice with Cole, where you have a scenario, you practice on your own, but I also provide a way to solve it. Then Practice on your Own, where you also go through a scenario, but without a set solution; which is the type of exercise that fit well into classes and for university instructors. You can pull exercises from the book or model your own scenarios based on them. Then you have Practice at Work, because one of the biggest feedbacks I was getting from workshops was: “This is great, I understand the concepts, but how do I actually translate that into my day-to-day?” This section is focused on that: you face a project, how do you break it into its component pieces, and get traction, get the right people paying attention, ask yourself the right questions, collect feedback from the right people at the right time.
The first book is very cut-and-dry: you found this, therefore you do that. It’s good for teaching, but in the real world, things are stickier, messier, there are corner cases.
The book’s been out only a couple of weeks, but some of the feedback or success stories I have were, well actually there was a tweet from Neil Richards this morning:
Some teachers have reached out to tell me: “we’ve already started the semester, so it’s too late to list it, but I’ve already pulled exercises from the book, and we’ll integrate it more fully into the curriculum going forward!” So far, it’s been really positive from an adoption standpoint. Which is something I was nervous about, right? Because it’s very different from the first book, and a very different book from what I’m aware of being out there.
CS: It is very unique!
CNK: It asks you to work, which is not what you usually expect from a book, right?
CS: Was writing any of the three parts composing the structure of the book a bigger challenge for you?
CNK: It was a lot of work! An interesting challenge. One of the benefits of the work we do is we get exposed to so many cases, businesses, industries, topics; the underlying challenges that people are facing are pretty consistent across domains. That means that we had a lot of examples to pull from. But you can’t just use a client’s example, right? So a lot of the work I got into was: how do I take this example and show it in a way that prevents sharing confidential or proprietary information, but allows people to practice the same sort of things as the original one did. For example, there’s one scenario that originally was a medical center looking at vaccination rates across their different facilities, and it turned into “numbers of cars sold per dealership” or something like that. It was interesting to swap things around where they no longer match with the original scenario, but from which specific lessons can still be extracted.
CS: Isn’t this a skill that we as practitioners should also build: to be able to look at an example from a field, and turn it into elements we can apply to our own projects?
CNK: Yes, and I believe it is an opportunity for people to get out of their comfort zone and really try something new. We often learn most from what’s outside our area of expertise. When we work with something familiar, we’re bound to overlook things that look obvious to us because we see it so often, so we forget that they might not be obvious to everybody. When we work with something outside our domain, unfamiliar metrics, it takes you out of this failure of looking through the same lens your audience does. It helps us focus on what are the things helping or prohibiting this alignment from happening.
CS: Alexandra, a member of the DVS, mentioned that you often emphasize the fact we have to keep our audience in mind in every project. How would you balance your work for an audience with varying levels of domain expertise or graphicacy?
CNK: Mixed audiences are hard, right? In an ideal scenario, we can communicate with a single person, understand their needs, what motivates them, and communicate for them specifically. Rarely is that the case in reality! We often end up having to communicate with groups made-up of various people, and stakeholders with different objectives. Some things that might help you in these scenarios:
- Challenge the assumption you have to communicate to everybody all at once. Sometimes there can be value, particularly if their needs are sufficiently different, to communicate to them in their way, to be more tailored to their needs.
- If you can’t do that, think of how you can still turn this into a successful scenario. So if you’re live with your audience, you can use verbal cues to mark that: “marketing people, you’re going to want to listen to this bit!” If you’re not live, then you need to think about how you are structuring your communication so that people can easily find and turn to the pieces that are the most relevant to them.
Another case where mixed audiences are difficult is when they have different requirements for level of detail: one audience member wants to get into the story, and another wants to view it line by line. You cant do this all at once, you’re dissatisfying somebody in any event! You might be able to identify these audience members that will want that level of detail, sit with them ahead of time so that when you get to the big meeting, you can focus on the big picture. Or plan some time with them afterwards.
Any amount of time and thought people give to: “how do I make this scenario successful for me?” will be well spent. Who is our audience? How are we presenting to them? What are we presenting in the first place? All of these things play into that. You may be looking for a magical recipe for success, but there isn’t one! The constraints are different every single time, and this is one of the reasons this work is so much fun. And there are a lot of approaches that could work. By being thoughtful about how we approach this, how we design our data, how we present it, we maximize our chances of success.
CS: This directly links to one of the tools in your book, the Big Idea sheet, where you go from all your possible audiences and narrow it down to one single target, doesn’t it?
CNK: True! Often times we have a big audience, but there’s someone or a small group of people within that that will be our primary target: decision makers, influencers, so we can keep them first and foremost in mind to be a bit strategic.
CS: Your first book came out in 2015; how has your craft and maybe your perspective evolved between your two books?
CNK: For me, one of the biggest changes when I think about where things were at in 2015 and how things are now, well, the fundamentals haven’t changed. There is a flow between the books since the process stays the same. Story, however, has evolved a lot since the first book, and with it components that are in Let’s Practice!: the narrative arc, tension … In reflecting back to where I was in my head about story with the first book, I was still trying to figure out how to talk about it and teach it in a business setting. And that’s one of the elements that has come a long way since then. Both in my own understanding of story, and how to be strategic and tactical about how to use it in a work-related setting or when you’re teaching about data.
I think it’s very interesting how people think about story: they imagine marketing, fluffy sort of stuff, which is not the case at all or doesn’t have to be the case at all. You can be very strategic about how you use story to really get people to pay attention to your data and focused on the right thing or hopefully the right action you are leading to.
CS: Speaking of story, you recently interviewed Nancy Duarte on the storytelling with data podcast for her book Data Story. She brought forward the concept of data “explainers.” What’s your perspective on this role (which I feel is growing) and how we can improve in this aspect of our practice?
CNK: I think it’s an interesting evolution. Not so long ago, technical skills were in high demand in relation to reporting and data visualization. But there’s a caveat to that: if you can make sense of the data yourself, but can’t help others make sense of your findings, a lot of your work’s value gets lost. It’s a hard skillset to develop though, because a lot of times people trained on the technical side have not spent so much time on communicating. You have this expertise, but can’t really take a step back to step into your audience’s shoes.
The people who recognize that there is value in this part of the process and invest time in honing these skills will benefit hugely. There’s an incredible amount of value to extract from work that is already being done, but not being communicated properly. Highly technical people, statisticians, who take time to invest in these so-called “soft skills”, how you talk about your data, how you put it into a business context so that they can thrive from your work, will really be rewarded.
There’s an incredible amount of value to extract from work that is already being done, but not being communicated properly.
CS: Would you say this is the missing link to be able to prove data analysis can bring value to businesses and stakeholders? Because they often think of work in terms of ROI?
CNK: ROI is always a difficult thing to measure. It’s rare that data visualization leads to an immediate, measurable action: sales went up by X percent, for instance. For me the efficacy test is more “if you needed something to happen as a result of your analysis, did this thing happen?” A discussion that needed to take place, a decision that had to be made — would it have played out differently had you not communicated it?
It is hard to measure, because the answer is most often nuanced, but it’s how you make all the data, how useful and used does it become?
CS: I had a discussion earlier this week with another member of the DVS, Francis Gagnon, who works more on the design side of communicating with data — versus the technical side — who said that often his clients reach out when they’re faced with do-or-die situations. “If I don’t get my point across, our project may fail or we may even lose our client,” for example. Being heard as a designer usually becomes easier in these situations, because they spend money on your expertise for a set project or amount of time.
Unfortunately, we’re not always in such a setting. How would you convince stakeholders that it is worth investing time and resources to build stories that will lead to effective outcomes?
CNK: You mean, how do you make the case that it’s worth spending time on the storytelling part? First off, you have to be conscious that it doesn’t necessarily mean you have to go back and entirely redesign a report, for instance. One way of taking it to the next level is to start with the report as it is, you go through it. Whichever type of report you have (dashboard, scorecard, spreadsheet), it’s a tool to display your data and see where things are working as they should, where are they less in line with our expectations… you don’t try to make a story out of every one of them. Sometimes the only story will be: everything is as expected.
It’s when it’s not that you can make the case and use a lot of what we do in workshops to reach out to your audience, direct their attention, tell them: “we can keep doing things the way we have in the past, and here is what’s likely going to happen. Or we could use the analysis we’ve already done, and here are the most important elements you need to pay attention to for us to reach the next level”. This is where your storytelling techniques will come into play. And as you start doing that, you will both build confidence in your skills, as well as credibility towards clients and stakeholders. They’ll know that you are focusing on the right things, and you’ll be able to lead them away from all the details to now form conversations around story.
To me that’s one of the most interesting things that happens when communicating with data effectively: the conversation shifts. It is no longer about the graph, or what it means, or requests for more data, more data, more data! It becomes: what does this data actually mean, and how can we use it in terms of our business? That is a magical transformation. It is never easy, it takes time, and it can easily become overwhelming. Start small, with low-hanging fruits, things that nobody is going to cry out after if they change and build a momentum there. You’ll be able to build the credibility to lead bigger changes down the line.
To me that’s one of the most interesting things that happens when communicating with data effectively: the conversation shifts.
CS: Listening to podcasts and interviews lately, and with what you just mentioned, it feels reassuring to know that it always takes time. That even people I look up to in the field can still ask themselves: am I doing this right?
CNK: Yes, and you have to become comfortable with trial and error, but learn something every time: learning what went well, what to emulate in the future, and what didn’t work for me and should be taken out of the way. There are no experts, everyone can get decent, increasingly nuanced in their approach to data, and it’s one of the reasons why this space is so much fun!
CS: Now that the book is out, I am assuming you have SO MUCH more time on your hands :) It may be kind of a left-field question, but has it freed you up to check some items on your reading list?
CNK: The book sitting on my nightstand is Elefant by Martin Suter, a Swiss author. I started reading it when traveling with my family in Zurich over the summer. It’s in German, which I understand but it takes really concentrating for me to follow everything. I’ve been reading a short chapter here and there, but plan to finish it in December when I take some time off work. All my other time goes to my family!
CS: Any chance you can give me a sneak peek of the next SWD Challenge? (which is now available here.)
CNK: Let’s say this: it implies going back to basics, but with a twist … Stay tuned!
CS: I have one last question (which Cole was kind enough to answer on a Twitter DM since I forgot it live, because, duh, I was chatting with Cole Nussbaumer Knaflic!!) There are a lot more hand-drawn illustrations this time around, with great work from Catherine Madden. (Your blog post on creating the book explains why you used this approach).
I, for one, do not always take time to sketch in my daily work; I suspect many people with tech backgrounds may also go straight to their laptop to prototype. How would you convince professionals (and stakeholders) to dedicate some time to sketching and low-tech prototyping in their design process? In other words, what added-value can sketching provide us?
CNK: While it seems more efficient to jump straight into our tools, there are important benefits of resisting that urge and beginning to design our data communications in a low tech fashion. This is best done after you’ve spent some time to get to know your data but before you spend time creating the beautiful versions of the graphs or other content.
Get a blank piece of paper and start drawing. Sketching has a couple of great benefits. It frees us up from the constraints of our tools (or what we know how to do in our tools). We’re also less likely to form attachment — I can sketch a crazy idea I have for a graph and see that it may not work. Whereas if I’d taken the time to create that same crazy graph in my tool I’d be less open to letting it go, even if it didn’t ultimately serve my purpose well. Finally, it doesn’t take long (big benefits when you spend just five to 10 minutes and the drawings can be rough and ugly and that’s totally fine) AND it streamlines the rest of your process.
Once you have a sketch of your graph or an outline of your story, then get stakeholder feedback to see if you’re on track or need to change direction. Possibly get feedback from someone else (or your stakeholder again) after you’ve made adjustments. Getting agreement up front means there’s less iterating in the rest of the process (still always some, which is inevitable, but this sort of low tech planning up front will make the rest of the process more efficient). There are a number of exercises in Let’s Practice! that will guide the reader to undertake low tech planning: forming a Big Idea, creating a storyboard, sketching graphs, etc., which from my perspective is some of the most important part of the process for communicating data effectively.
(Spoiler alert: this was the topic of Cole’s Tableau Conference presentation, see video below!)
Thank you again to Cole for taking the time to speak with me. Congratulations again on the new book and best wishes for all the other activities coming up at storytelling with data!
To find out more about Cole and storytelling with data’s activities, use the following links:
Special thanks to Alyssa Bell for her generous collaboration on this article!