When Data Gets Up Close and Personal

People love statistics, especially about themselves. How can we use this?

People love statistics, especially about themselves. With the rise of personal informatics we see the question “How am I doing?” getting ever more popular. I was wondering if we could use this to create a game out of email that would make life easier and happier at the same time.I’ve recently become interested in sites that answer the question:
“How am I doing?” Apparently so have a few other people…

New startups offer to help individuals improve their performance— in a variety of areas: Rypple helps individuals improve themselves professionally through peer feedback. Klout helps people realize the effect and reach of their tweets. Dopplr lets us track our travels (and our carbon footprint). iPhone apps like trackyourhappiness.org produce a happiness report based on simple self-reporting. According to Hunch, I’m “an optimist” (based on answers to hundreds of microquestions). And, the app I’m currently working on is designed to help people follow through on commitments.

But what about business applications like email, time tracking or  business intelligence applications? Is there a place in this “buttoned-up” world for these same kinds of personal feedback loops?

A little perspective

If we step back from software and Web apps, the idea of tracking performance is nothing new. Baseball cards reveal every numeric detail of player’s performance. Schools grade students’ performance. And many people routinely track their diets and workouts. Even in an agricultural society, having a good or bad harvest is the ultimate comment on how well a farmer did that season. But technology is making it easier to passively monitor personal details we couldn’t monitor just a few decades ago.

Where games like Pac Man rewarded us with various fruits, more intricate games like Guitar Hero break down our every action: percentage of correct notes, longest streak, breakdown of success by parts of a song– the list goes on! Through this report, you quickly learn–or confirm–which sections you need to work more on (and you can find out just how close you are to being a rock deity). But those are games, right?

What about hypermiling? For the first time in my life, I’m conscientiously tracking my gas mileage. I’m really curious to see if my diesel car can attain the advertised 56 MPG (or at least something in the 40ish range). Thanks to a couple of iPhone apps, I can track any number of automotive details. This is also the first car I’ve owned that displays real-time MPG feedback. The result? I’m learning to adjust my street driving (slower starts, coasting at times) to improve overall gas mileage.

167 miles per gallon?! (Some coasting may have been involved...)

167 miles per gallon?! (Some coasting may have been involved...)

Is hypermiling a game? Does this really even matter?

“Serious Games”

What we’re really talking about is setting up systems whereby individuals can (1) see in a tangible way (2) reflect on, and (3) learn from their past behaviors. Think about what Mint is doing for personal finances, or Nike+ for workouts.

From report cards to Pac Man– we’re talking about the same thing: feedback loops that affect future performance.

Through personal informatics or competitive scoreboards, when someone (or something) holds up a mirror to our behaviors, we gain information to help us improve in that area. Consider twitter. I think it’s something as simple as twitter’s follower count (and now list count) that has made this “How am I doing?” concept top of mind for many people. I mentioned hypermiling earlier. And we’re seeing new devices to help us monitor energy usage in the home. But why aren’t these same kinds of feedback loops more common in everything from invoicing tools to email apps? Why doesn’t email have a scoreboard? Seem far fetched? We’re talking about individuals– we respond to the same psychological nudges, whether we’re working or playing.

Making a game out of email

“Okay,” you’re saying, “interesting ideas but how might we apply these ‘numeric nudges’ apply to a more serious subject?

To test my hypothesis– that yes, “serious” applications can be playful– let’s go through the process I’d use to transform email. Starting with the stated goal of “inbox zero,” let’s design a system that might help us become better email players. Let’s create The Game of Email.

1. Identify specific behavior patterns to encourage (or discourage)

What behaviors do you want to change? Is there a behavior you want to see more? Maybe it’s a bad habit you’d like to eliminate? For some applications, identifying this can be challenging. Fortunately with our example, thousands of bloggers and self-help gurus have already identified both good and bad email habits and translated these into helpful tips like:

  • Never open an email twice;
  • Read emails in the order they were received;
  • Answer briefly;
  • Respond in a timely manner;
  • Only check email twice a day (or once an hour);
  • …and so on.

We’ll build on these tips for our little email game.

2. Translate desired behavior patterns into data that can be passively tracked and measured

Once you’ve identified (and prioritized) specific behaviors to influence, let’s see if these can translate into specific data that can be passively monitored by the system. Sometimes, translating an idea is straightforward: With something like “respond in a timely manner” we can start a timer from either the moment an email is received or opened for the first time. Be careful though, depending on your context, even a simple translation like this may need to factor in things like business days or adjusting what “timely manner” means to various people.

gmail shown with countdown timer

You don

The biggest challenge you’ll find is with tracking qualitative behaviors. For example, if I wanted to know how clearly I’m communicating in my emails– that is much more difficult to measure compared to simply tracking if or when I responded to an email. That said, tracking qualitative achievement is possible when we introduce a social layer, something I’ll mention below.

3. Attach points to these behaviors
Now that we’ve identified the behaviors we want to encourage, and determined how to track those behaviors, let’s recognize them. We’re going to award and deduct points for the behaviors listed above; points earned will form the basis of our scoring system.

Using “never open an email twice” as an example, a good system would encourage you to:

  • Read and delete;
  • Read and respond;
  • Read and file, or;
  • discard (without reading).

So, we offer helpful text to remind people of these 4 options (think of this message as training wheels that eventually go away) and then we award points based on behavior:

  • +10 points for taking action when you open email;
  • 0 points for opening an email a second time;
  • – 5 points for opening it a 3rd time (and so on).

If we were to create a similar point scale for the other behaviors we identified, you end up with a total number of possible points and actual points earned. So, for a specific email exchange, you might get something like this:

4. Translate points into a periodic score and other useful information
While tracking performance per email exchange is kind of interesting the first few times, this novelty gets old quickly. Think about ways you can introduce cycles into the system– an end of month reports is a great idea to see how we’re doing overall. By slicing the data in different ways, we could also learn what our average response time is to specific individuals. Or maybe we might learn what time of day (or night) is best for us to respond to emails. The possible types of reports are limited only by what you can imagine: just don’t lose sight of what is interesting and worthwhile to people.

5. Display the score in a fun way
Far too often, we have interesting data, but it’s not displayed in a way that is compelling. In this discussion of numbers and data, it’s easy to lose site of the fact that we are emotional beings; there’s a growing body of research exploring how our affect governs everything from decision making to memory. Make the time to look at the what you are revealing and determine if there is a more compelling or emotional way to present that information. Dopplr chose to represent “personal velocity” (distance traveled in a year) not as a number, but as animal that moves at approximately the same velocity. What if credit scores where represented as a hot air balloon? Or a measure of collaboration might be represented as a bee hive? Get creative with how you represent the data– our brains will thank you for that with extra attention.

6. Create Rules to Translate Data into Helpful Information
While cumulative scores and fun representations are somewhat useful, think about how you can turn specific activity patterns into helpful tips. This can be a tedious process of defining the rules and correlating messages, but the resulting personalized tips can be quite helpful:

“Ouch! You only responded to 38% of your emails in a timely fashion. This may be due to your lengthy (avg 17.4 sentences) replies. For next month, focus on shorter responses in a shorter timeframe.”

If our “score” hinted at a story, the story is explained in this text.

7. Set Challenges
Competing with oneself is a powerful motivator– if you provide something to compete against. This can be a best “winning streak,” a “top score,” the allure of “the next level” of mastery– the possibilities are endless. A boring game is one that doesn’t offer me ever increasing challenges. While mastery of email may be a challenge for some, up the ante for those who are at 100%. Decrease the time allowed to respond to incoming emails or introducing new barriers– you can no longer check emails as frequently as you did.

One interesting note here, while many games are built on a system that encourages you to “level up” as you increase experience (think Karate belts and most videogames, etc.), this assumes a beginner level. In business contexts I’ve found the analogy of a “credit score” to be more often the case; that is, you want to maintain the highest level possible, which you may or may not have been when the “game” began.

8. Add social cues
While competing against your own best score is a powerful motivator, social cues are much more powerful. “How do I compare to my peers?” What’s an “avalanche” for other people? 30 emails? 500 emails? How does their situation compare to mine and are they any better?

Imagine disclosing your “email ninja” score with others (and likewise). While there has been some heated debate about the use and abuse of leaderboards, we do like to know how we’re performing relative to other friends who are also “playing the game.” In a somewhat quantified way (or maybe something more playful, like Dopplr’s personal velocity), you could discover those folks in a similar situation to you who seem to have their act together. Maybe a “privilege” of the game is earning the status to share a tip or two with others in your “game of email” network.

9. Have fun. Make it interesting
This is a catch all for all the other fun things we haven’t even discussed! Think about pleasant surprises. Or ways to create scarcity (Attent™ with Serios™ has introduced scarcity by creating a virtual economy where you earn and spend points to increase the importance of an outbound email). In a previous article, I discussed how curiosity can motivate people to action— how could we arouse some curiosity here? What about earning privileges, such as new features or the customization opportunities that tap into our desire for self-expression. What about exchanging virtual gifts? What about injecting some humorous language…


We could go on specifying the details of The Game of Email. But you get the idea. We’re introducing feedback loops that tell me — in objective terms — how I’m performing. Not just at periodic intervals, but along the way with tight feedback loops that indicate how I’m doing now, so I can adjust and change course.

The rules are up to us. The game can be personal or social. We can layer on tons of other game mechanics (challenges, levels, variable rewards, prizes and what have you). But underneath it all, there’s a system offering me reflection on my behaviors.

Could business applications benefit from this kind of thinking? Would these numbers change anything? It’s a nudge in the right direction.

Stephen Anderson

Stephen recently published the Mental Notes card deck to help product teams apply psychology to interaction design. Between public speaking and consulting, he offers workshops to help businesses design fun, playful and effective online experiences. He’s currently writing a book about “seductive interaction design” that will be published in 2011.

16 comments on this article

  1. Great article, Stephen. And a lot more fun than Mark Hurst’s slightly pedantic “Bit Literacy” approach. (see http://www.goodexperience.com)

    My problem is that folks who write me are not particularly disciplined; they combine several subjects in one e-mail. So even if I can dispose of one issue, I have to file the e-mail for future reference regarding the other couple of items. And when I cannot file things (even if I’ve answered), this stuff ends up filling up my inbox. Case in point:

    “Hi Eric,

    Thanks for the Powerpoint slides. Can I reuse the visual on slide 23 if I promise to credit you?

    BTW, I’ll also be at the Hyatt in Savannah. Remember to bring the plastic duck you promised my daughter.

    I can’t remember if you have my mobile number: +212 555 1212.”

    Yikes. Stuff to remember, stuff to respond to, stuff to enter into my Outlook (and transfer to my phone). Not an easy e-mail to deal with – especially if I’m rushed.

    Any suggestions?

  2. Pingback: Most Tweeted Articles by Designer Experts

  3. Ha! That is a problem– How do we help other people change their behaviors!

    IF you can get other people to “play the game,” the solution is a lot easier– you simply ding them for breaking the rules! Of course, the bigger game we’re playing is one of social norms and email etiquette. It’d be nice to have a polite way of making people aware of some good, best practices. Perhaps through modeling good behavior yourself and hoping for some positive mimicry? On a related note, have you seen http://five.sentenc.es/

    But, until we change other people’s behavior…

    Within the game I describe above might be a fifth option for “needs longer response,” where the things become a task item moved to a separate folder (possibly with a different set of rules and objectives?).

    Since my email example was for illustrative purposes, I’d love to hear from others how you might apply the approach I describe to solve this problem…

  4. Ton Bil on

    Ooh, lovely thinking for me, thank you. (found via @IATV)
    It’s lovely because I will want to use this playful mode when Workpatch (my online + mobile web service) is going to feedback to users how they’re doing.
    Workpatch is about matching supply and demand of odd jobs, gigs and so on. And here you’ll find things like evaluation, exploring talent, interaction with neighbours, etc. The cues and perks that you come up with give me good inspiration. Thx! (Oh, sorry, my blog is in Dutch).

  5. Arvind on

    that’s an interesting collection of design principles..

    however, i do think that your email example is a bit misguided and betrays a lack of understanding of how email is used in reality. why do i say this?

    email is a multi-purpose communication system: many uses intersect in the same medium, leading to many different usage behaviours. any personal tracking system that attempts to normalize behaviours is going to essentially wreak havoc, because it will evaluate legitimate behaviours (referencing an email over and over again as you work through a related task) identically with ‘undesirable’ (i use this word advisedly) tasks (reading and ignoring an email repeatedly).

    secondly, your solutions completely ignore context, which, as any cursory examination of the computer-supported collaborative work (CSCW) or HCI literature to date will show, is central to most cognitive behaviours related to work. that is, actions that are desirable in one context (letting oneself be interrupted by an email about an important project) may be ‘undesirable’ in another (letting oneself be interrupted by a joke email while in a meeting). unfortunately, however, current technology is crap at understanding context – mostly because no corp has taken a serious stab at it – so dealing with (high volumes of) email is largely a brutish experience.

    so, we should really try to fix the technology, not people. beware of normativity as a default response!!

  6. Arvind on

    i should probably add that i’m not critiquing your email example per se, but to point out that the same problems exist with any socio-technical system, and thus need to be kept in mind when designing any behaviour influence/modification systems. (for real life examples, think about road signage and Dutch road safety engineer Hans Monderman’s insightful responses; http://www.wired.com/wired/archive/12.12/traffic.html)

  7. Hi Arvind!

    Excellent comments!

    You’re absolutely right about email being a multi-purpose communication system. How do you determine evaluation criteria when usage is so different? In an earlier draft of this article (one overly focused on the game and not the principles), I did present the idea email archetypes. I do think after more than a decade of email usage, we could arrive at a set of archetypal email interactions (mailing lists, direct emails, system notifications, etc.). If we can agree on some common email patterns, and assist users with identifying these early on, I do think we could define normative behaviors for specific types of exchanges. Good observation!

    Your comment about context seems to me the more challenging one– and a critical consideration. I’ll offer up a few thoughts:

    – Technology context might indicate social norms: Many people use Outlook at work (for business), and gmail or Yahoo! mail for personal use. These platforms can be indicators of context. In my case, I use one platform for professional and personal use– but this reflects how blurred these lines are for me…

    – Are there desirable behaviors that a dumb system can help enforce? For example, should you be checking email during a meeting? Should you allow yourself to be interrupted by email throughout the day? Increasingly, many of the best recommendations for email use converge on similar ideas about setting rules for distractions. I might be in a situation where it’s good to check email every 1-2 hours. For others, it might be best to check twice daily. I think the game described here could help you choose a good default based on your needs/goals…

    – Social context: One thing I mention in passing in how we evaluate performance in a system where usage does vary so much. For one person, 10 emails a day is the norm. For someone else, juggling several 100 emails a day may present no problem. Or juggling 10 or 200 could be a daunting task for another person. As with exercise programs, there are standard patterns of what constitutes “good,” but we don’t hold everyone to the same standards. For my BMI, I may be doing great to run twice a week for half an hour. For someone else, that could be major backsliding. What I think a system like this would offer is a chance to look at people within my social circle operating under similar conditions– I could see how I’m doing relative to everyone, but analogous to users who share similar circumstances.

    I agree we should look for every opportunity to fix technology deficits before making users change their behavior. What I’m trying to describe is how we might go about creating a system whereby people can realize their goals. This may come down to whether we view email as a tool to be used or a platform within which to operate.

    By the way, that’s an excellent article you mention! Precisely the kinds of things I love learning about…. I certainly favor reduction in the UI wherever possible. My article indicates an additive process, as I’m thinking about how this could be realistically achieved as a plugin to extend the email platforms we already use. If I was designing a new email platform from the ground up with these things built in, the execution would probably be different– I’d build some of these ideas into the architecture of the system– take a “break away from the conventional ideas that have got us in this mess, ” to quote the article. But, that would introduce a bigger problem: asking people to change their email platform (which is a much, much more daunting challenge than the gaming I describe in the article!). Another thing I love about the Wired article, it mentions “bettering the quality” of life as a motivation for redesigning the roads. While I make no mention on this in my article, I’d want to eventually included system that rules that help you make better (for you personally) decisions–it’s not all about inbox zero or efficient responses. In the end, email is just one tool (or system) to facilitate communications. Viewed from this perspective, there’s plenty more we could do to re-imagine not only email, but the various ways we interact with people and entities…

  8. Pingback: Recent reads - using data « learn.amniisia.com

  9. don gale on

    This sounds really interesting. Is it something that’s available, or are you speaking from a developmental standpoint?

  10. Pingback: Weekly Roundup: Design Related Links #5 « Discovery Session… by Gerard Dolan

  11. Pingback: uberVU - social comments

  12. Pingback: A game of email

  13. Pingback: » The Web and Beyond report – day 1 Johnny Holland – It's all about interaction » Blog Archive

  14. Pingback: A propos de Tweet Runner « Florent Deloison | blog