EPISODE 13
goals like Google

TIMESTAMPS

00:32 | Intro
04:48 | OKRs
16:41 | Listener Q&A
28:10 | Updates 

TRANSCRIPT

Welcome to storytelling with data, the podcast where listeners around the world learn to be better storytellers and presenters with best selling author, speaker, and workshop guru Cole Nussbaumer Knaflic. We'll cover a wide range of topics that will help you effectively show and tell your data stories. So get ready to separate yourself from the mess of 3D exploding pie charts and deliver knockout presentations. And with that here's Cole.

00:32 | INTRO

Happy new year! Thanks for tuning in. I’m going to get a little personal  to kick us off today. Actually, I’ll start with a bit of trivia—you may know this, but: my husband, Randy, and I both worked at Google. I worked at headquarters in Mountain View and he was based in Zurich, Switzerland—that’s actually where we met, we did the long distance thing for a while, had some lovely European meet ups… ok, now I’m probably getting a little too personal.

So, we both worked at Google: one thing that Randy and I each experienced there—it’s like a religion, really—is the quarterly goal setting process: OKRS - Objectives and Key Results. I actually attribute a huge part of SWD’s success to this, which I’ll talk about more soon. And—perhaps surprisingly—I’d say OKRs have contributed to my marital success as well. It was relatively early on, I don’t remember exactly when, but I was definitely already using them on the work front with SWD, when Randy suggested that we set OKRs as a couple… personal OKRs.

Now, that might seem a little strange, but it’s actually been amazing. We sit down together—usually after dinner and dishes are done, kids are in bed and generally with a glass of wine in hand—and set our goals for the quarter. Perhaps stranger than that, at the end of each quarter, we do this again and discuss and grade ourselves on how well or poorly we achieved each of our stated objectives. That’s actually an equally important part of the process as setting the goals in the first place. Right, this idea of coming back and reflecting to say, well, what did I actually accomplish relative to what I set out to do?

I’ll be honest: while I consistently set OKRs on the business side, we sort of fall into and out of the habit on the personal front—we haven’t always been consistent. Though stepping back, I feel comfortable saying that we’ve been more in sync...possibly even happier when we have gone through this process regularly together. There is something about setting—actually articulating, agreeing on, and writing down common goals that, for me at least, helps me to make my actions more intentional. There’s I’m sure also something about the accountability aspect—both that I know someone else is in the picture who is expecting I do what I said I’d do, as well as accountability to myself because I know we’ll sit down and grade them at the end of the quarter and I’m a perfectionist, I want good grades!

Our OKRs have run the gamut over time, from things at home, like how often Randy will assume dinner cooking responsibilities, to stuff with our kids, like making sure all three of them get to the dentist for a cleaning, or one that was amusing as I was scanning through some past ones was “Randy sing and plays Somewhere Over the Rainbow on the ukulele for Avery”...to stuff with us as partners, like how many date nights we’ll have or, let’s just say some more personal stuff that I won’t get into!

It’s a new year, right, 2019 just started. This is the time of year when people often set resolutions. I don’t like resolutions. You resolve to do something: I’m going to lose 20 pounds or take up golf, I don’t know. But rarely is much thought given to how (beyond joining the gym or the club), and even rarer does anyone ever check back in a reasonable timeframe to assess: did I do it? And rate their success. That’s why I love the OKR process so much: these latter pieces are built in. 

04:48 | OKRs

So let’s talk about OKRs!

OKRs. What are they? So I mentioned this earlier, but OKR stands for objectives and key results. So they're made up two components really; your objective, which is the goal. This is what you want to accomplish. And it should be significant and communicate action. And then the second component are the key results. These are how you'll accomplish your stated objective. They should be aggressive, but realistic. It's very important that they be measurable because at the end of the quarter you're going to come back and actually assess whether or to what extent you completed each. We'll talk about this more momentarily.

Oftentimes, there is some sort of time component included. This may be a deadline or a certain frequency with which you're going to do something. And then they're limited in number. You typically will only have two or three key results that are supporting a given objective. The objectives themselves are also limited, where you'll have somewhere on the order of three to five objectives typically for a given quarter. There is a great Google ventures video describing OKRs that I'll make sure we link to in the show notes. 

Let's talk through an example. So an example OKR for me this quarter, I'm going to be taking my first strip to Australia and New Zealand, this will be my first name doing work there, so I might have an objective around regionally targeting business development. Could be something simple like “grow storytelling with data business in Australia and New Zealand.” I want to make sure I have a successfully expansion there. 

Then my key results could be something like register at least 30 people in each of the three workshops by the middle of January. By the way, if you're in Australia, New Zealand, or the South Pacific, you can help me achieve these key results by signing up today at storytellingwithdata.com/public-workshops. Another key result could be to book work with two local clients for the trip. Or I might have one present in three forums to share storytelling with data with the broader community and create future potential interest. These would be things like meet-ups or Tableau user groups or university guest lectures. 

At the end of the quarter I can look back and assess to what extent did I accomplish each of these key results. Then together I can say well, if I accomplish my key results, then the objective has been met, or figure out to what extent it has been met. And then that helps me both assess progress, right, to see am I doing the things I set out to do, but then also it helps me plan for the future.

Now, if you're working with others, it can sometimes make sense—or often makes sense, I should say—to have common objectives, where you might share and multiple individuals have the same objective, but potentially different key results for accomplishing that objective. So if I take another storytelling with data example, Elizabeth is on the team here at storytelling with data. If we look back in time, at one point we were trying to get comfortable, both her and me, with her presenting trainings and workshops to client organizations. And so at one point we had an OKR or an objective that said “Cole and Elizabeth each agree that Elizabeth is ready to present on her own to clients” in the given quarter. Then we each had different key results that would help us individually, but also together, as it makes sense, to meet that objective.

One thing that's been interesting for me over time in the way that I use OKRs, right, you think about setting goals of things you want to accomplish in a positive way. I've also sort of used them in a negative way. Maybe that's not quite the right wording. Something that has been a constant struggle for me historically and, you know, this is maybe the case for anyone who owns and operates a business, but there's this desire to want to say yes to everything. Because you never know when things will change and so there's a desire to want to, what is the saying, like bird in hand. Right, you want to take advantage of that. But you can't actually do that if you want to be successful because you need to be focused and say no to a ton of things so that you can be focused.

One way personally that I found OKRs to be really effective is to set OKRs about things I'm not going to do. So historically, I've had OKRs around things like no more than X trips in a given quarter or no more than Y consecutive nights away, because there's this desire, when the inquiry comes in, to want to figure out how to make it work. Well, I'm going to be in New York, so I could squeeze in one more thing. But when I know that I'm going to grade this objective at the end of the quarter that I have that says don't book any more than, you know, a certain amount, then it's easier for me to say no, because I know that means that I'm actually helping myself achieve a goal when I do that.

So I've mentioned grading, but this is a super important aspect of the overall process where at the beginning of the quarter—and this is usually in the first week or two of the quarter—you set your OKRs for the quarter. Often you’ll get feedback from someone else, a manager, for example. Randy, since he's familiar with the process, he's my accountability check, he's the one who gives me feedback to say no, that's not aggressive enough, or that's not really measurable, can you refine that. And then at the end of each quarter you grade your OKRs. And this is going back to those key results and assessing. So I use a score on a zero to ten scale where zero means no progress at all. All right, and this happens sometimes, maybe you'll reprioritize or miss something. And then ten is you accomplished what you set out to do to the fullest. Right, a hundred percent. And then you can have scores in between that as well, which is often the case. 

You assign a score for each of the key results and then you average those to get the overall score for your objective. And then if you want a summary for the quarter, you can average. I typically don't recommend averaging averages, but in this case I think it's probably okay as a signal to average the objective scores to get an overall numeric for the quarter. And for me the actual number doesn't matter so much. And I'm pretty aggressive in how I set my goals, so my overalls tend to be in the six to eight point range. But it's a useful—less of a useful data point, but more of a useful process. Right, again this idea of not only setting out to do something, but then reflecting and learning from that process. So, for example, I'll typically have a few places where I score zero, and then I want to pause and reflect and say well, did priorities change on purpose or did I actually just not do or devote time or make a priority something that I actually probably should have done? So then that changes how I think about it for the next quarter.

And getting feedback is important as well. Right, somebody else to help you really be true in your assessment. And I mentioned for me that's Randy because he's accessible, he knows the process, he knows the business, and so he can read through and say, you know, yeah, that makes sense to me. Sometimes he'll challenge me. Oftentimes the places he challenges me are actually hey, you're being too hard on yourself here, bump that up a little bit; really you did achieve that. We have some really good conversations that I think are useful for sure as part of the process as well.

When it comes to cadence, so quarterly was what I did when I was at Google. I think the performance process there is moved to semiannually in some cases. I'm not sure if the OKR process has followed that or if it's still consistently quarterly across the organization. But for my business, quarterly feels like the right cadence. Long enough that you can accomplish meaty things, but short enough so that if you go off track or aren't prioritizing in the right way, you get relatively quick feedback on that when it comes to the reflective piece. And especially for businesses that are changing, I think quarterly is a good cadence. Though I have found over time and, again, early on this wouldn't maybe have even made sense because I wasn't able to step back and think farther out, but over time, I’ve found it useful to also have a bigger goal-setting process where near the end of each year I'll step back and do strategic planning for the coming year. So we just did this process for 2019 where I got really specific in terms of numeric targets for 2019, main areas of focus. Be clear on those, and then the OKRs can break some of those main areas of focus into sub pieces that can be accomplished within a given quarter. So these two dovetail quite nicely.

The overall message here is really thinking about how you can use goals to be intentional in the way you work. I've had good success and highly recommend this quarterly OKR setting process and grading process that I've talked about. But I think it's worthwhile to consider how can you use goals to be more intentional in all aspects of your work. You could think about setting goals for a given project at the onset. Maybe it's a goal around having fewer iterations or incorporating storyboarding into your planning process or working with someone new or avoiding filler words when you present. Or maybe even for a given graph you have goals that you set and articulate and assess. This could be things like I want my graph to be so accessible that we don't spend time raising and asking questions about the graph. Rather, me and my audience are able to jump straight into the data and talk about what can we learn from it or what actions can we take on it. Or maybe there's a goal around being strategic in how you use color to direct attention. Or perhaps you have a goal around using a new tool. Actually, on that note, the current storytelling with data challenge focuses on doing just that; using a tool that is unfamiliar to you. That challenge runs through the 15th of January, 2019. You can find more information at storytellingwithdata.com/swdchallenge.

So no matter how you're setting goals, I recommend step back and be specific in the steps that you'll take to accomplish them. Consider what does success look like, how you can measure to what extent you've accomplished the goals that you've set, who is going to help hold you accountable, and, finally, when and how are you going to assess. In this way you can use goals as a way to push yourself forward and be intentional in the way that you work.

16:41 | Listener Q&A

Next, let’s shift to listener Q&A

Gemma: Hi Cole, I'm loving your book and the podcasts. I look after the Analytics Team in a small business and we're working to improve our data visualisations learning from your examples and recommendations. The way we work is our team develop the data aspects of the presentation and we know the story, but this is usually presented to clients by a colleague in the Client Relationship team. How would you recommend we help bridge this gap if we're not the storyteller in the live presentation?

Awesome question, and, yes, this definitely complicates the challenge when you are preparing data stories that someone else will present. This means your audience is actually two fold: those who are presenting your work and the audience to whom they are in turn presenting. In these cases, it is especially important to use color sparingly and words wisely. Color sparingly to help direct the eventual audience's attention where to look, and words written physically on the document making it clear why you want them to look there. That way there’s no confusion, either by your immediate audience of the Client Relationship team, or of the eventual audience to whom they are presenting. Now, if you're using any sort of slideware (PPT, Keynote, Prezi), you can make use of the notes area for additional support for those who will be presenting, using this to make the overall point or takeaway clear (though hopefully the slide itself does that already!), add context or answers to questions that may arise—this helps ensure the person who is presenting is prepared—that way they have all of the info but it doesn't all have to be on the slide directly.

I also highly recommend—if at all possible—getting together with the person or people who will be presenting the material you develop, both before and after. Before so that you're on the same page as to what's important, relevant context they may need, answers to anticipated questions, and so forth. And then afterwards, to debrief on what worked and what didn't so that everyone can continue to refine for next time. I know that this isn't always possible, but if it is, this can be super valuable for all involved!

Paul: I’m reading through your book Storytelling with data, and on page 18 of the Kindle book it says that I can download the excel file examples from your website. I’ve looked on the site and can’t find the files. Could you please point me to the right location? 

This is an easy one: all of the Excel files with data and graphs from the book can be accessed and downloaded at storytellingwithdata.com/book/downloads. I welcome you to use those files to learn or to teach, with proper attribution, and you’ll find that detail when you follow the link, which I’ll also put in our show notes.

Michael: Hi Cole, Thank you for all the great resources. I’m currently working my way through your podcast, blog, and book all in tandem. I’m interested in what tips you have for consumers of graphs and visuals as opposed to creators. I work mainly with teachers and school leaders who have a wide range of comfort working with and interpreting data. One of my goals is to increase data literacy and comfort across the schools where I work. Especially as I begin my new role, I am conscious of what data visualizations people have been exposed to (mainly bar graphs and pies from the state). I know part of my job is to make the information as easily interpretable as possible for my audience, but do you have any suggestions on effective resources to pass along or should my focus be on conducting training/workshops?

Awesome question. I've honestly not thought a ton specifically about resources for increasing graphical literacy, but agree this is absolutely something we can/should do. Elijah Meeks touched on this in the series he published on Medium about what charts do (though probably more of interest to you than to your audience). Alberto Cairo's upcoming book I think will be a great resource for both you and your audience—it's meant primarily for consumers of data viz—planned to be released in Q3.

I just did a quick scan of my bookshelves, but what is there is definitely more useful for the creators than the consumers. I do think your strategy of sticking with what is known is a good one—sometimes it can be useful to start there and then pivot from a known (grounding people there) into something less familiar (if there's value in doing so), so you've set the foundation for them, gotten everyone on the same page, and then hopefully brought them along with you to the new view. Also, I think especially in the scenario where you have people who have different levels of familiarity and comfort with data and graphs, that the "so what?" and story become really important—if these pieces are clear, the graphs become more like the back up rather than center stage. So for every graph you show, think about where you want your audience to look and take intentional steps to highlight and focus attention there, and also what you want them to take away and put that into words. That can make even a difficult or unfamiliar graph feel less intimidating.

Jen (Notre Dame) - 2 questions: My students just turned in a data viz redesign and several of them are flipping the X and Y axis to have time (years in decades) on the Y axis. I’ve always taught that time, which in this case is ordinal is conventionally on the X axis and we read left to right so this makes more intuitive sense. Further, when you have two variables where they are related (in this case years and public and private school enrollment) the result is that one is dependent on the other…so the dependent variable goes on the Y axis and the independent on the X axis. Since we aren’t controlling time here it’s the independent variable and thus goes on the X axis. Am I right? Anything I’m missing? What exceptions are there where time would be presented vertically? That just seems strange to me.

Thanks for your question, Jen, and yes, you're right that time is more intuitive on the x-axis going from left to right. Turns out we (in Western cultures) have pretty strong built in construct for this, so we can actually make it easier for our audience by orienting our data this way, which it sounds like is what you've been recommending. This makes sense.

That said, it's definitely not a hard and fast rule, and there will be exceptions. For example, say you work at a company that runs an employee survey each year and you want to focus on the latest year's numbers, but want a few past years there for comparison. Let's say we want our audience to see the latest year's data first. We wouldn't want time running backwards (starting with the latest year's data and then going backwards in time as we move left to right—that feels super confusing!), so moving time to the y-axis could actually be a good solution here, because of the way it would break us out of the time moves left to right expectation. With horizontal bars (for example), we could have the latest year's data at the top and then the historical data following (moving backwards in time as you go from top to bottom). Or I could imagine a similar set up with a dot plot.

Jen (Notre Dame, Q2): One thing that my students have a hard time with is humanizing the data. They aren’t really sure what that means, how to do it (other than ask "what does this mean for PEOPLE?”), and when I talk about Duarte’s work in Resonate (the hero’s journey and a likable main character who experiences a conflict) they think it sounds silly for business presentations. My finance MBAs really hate it and say there’s no place in their world for that fluffy stuff. They are tough. Any words of wisdom? 

I find this super interesting on the humanizing data piece. I used to encounter this a ton—the sentiment that "communications" are soft skills that technical people shouldn't bother with—I’d say, in the business world at least, that view is shifting. It's becoming clear that you can have amazing technical skills, but if you can't then communicate effectively to someone who doesn't work from that same skill set, you fail. Your work might be amazing, but if you can't communicate it clearly, then all of that work is for nothing!

In my workshops, I often use a personal story to illustrate. I'll tell it early on when I compare story to a list of facts on a slide. Later in the day, people don't remember any of the facts, but they are able to recall these super specific details simply because they were woven into my story. That's not to say that we should use personal stories to communicate, but rather to illustrate how the ability to understand, follow, and be able to repeat well-told stories is something that's hard wired in our brains, and we can use that to our advantage when crafting data stories as well (similar to how you want to generally work with the natural construct that time goes from left to right in a graph, using this natural construct of story can help make things easier for our audience—and when things are easy, they don't have to do work, they are more likely to pay attention and hopefully have a meaningful conversation or act in the way we want them to).

The way I teach this has shifted over time (and has definitely become more nuanced since I wrote the book, which is one of the reasons I'm working on some new content which will be available later this year). I focus now mainly on the narrative arc (plot, rising action, climax, falling action, resolution). For me, in a business setting, the order of the arc is less important (we can swap around the pieces as makes sense depending on the situation), the most important thing is the identification of tension. The tension is what contributes to the rising action and climax. This is not about making tension up, but rather considering what is the tension that already exists (if there weren't tension, I'd argue we don't have anything to communicate about in the first place). The tension is not what's at stake for us, but rather what is at stake for our audience. Going back to Duarte, who you mentioned before - and she makes this point quite clearly: you are not the hero in the story—the audience is. Then the resolution becomes how our audience can resolve the tension that we've highlighted. So people who think story is fluff when it comes to data likely need to get more nuanced in how they are thinking about it. My view is that story used well can help us get our audience’s attention, build credibility, and motivate them to act.

I think that’s a great point to end with. Big thanks to everyone who submitted questions. If you have a question, you can email it to askcole@storytellingwithdata.com.

28:10 | Updates 

Thanks very much for tuning in! Before we wrap, a couple of updates:

• The public workshop for the first half of the year has been set! This includes stateside sessions in Charlotte, San Francisco, New York, and Milwaukee, my first trip Down Under in March, with sessions in Auckland, Melbourne, and Sydney, and European workshops in London, Dublin, Copenhagen, and Zurich. Details and registration can be found at storytellingwithdata.com/public-workshops.

• The current #SWDchallenge is live and runs through Jan 15th: this time I encourage you to try out a tool for visualizing or communicating data that is unfamiliar to you. I think we’re all going to learn a ton from this one and we’ll share back a recap post later in the month that includes all examples created and commentary shared. Details at storytellingwithdata.com/swdchallenge (all one word). 

• Also a quick teaser for the next episode of the podcast: you’ll hear from a number of people whose names you’ll recognize about the varied paths they individually undertook to learn to visualize data. It’s shaping up to be a fascinating episode and I’m excited to soon share it with you! 

• Speaking of the podcast, If you like what you hear, please subscribe and share with a friend.

With that, be sure to follow @storywithdata on Twitter and Instagram. Check out all the great resources on the blog @storytellingwithdata.com. Thanks very much for listening!