I’ve been thinking a lot about user engagement lately, which might be why Nike+ Experiment article in last month’s Wired caught my attention. Although the article only briefly touches on the topic, there’s a lot that can be learned by studying very successful products. I’m going to head out on a brief tangent to take a look at three valuable lessons, then circle back at the end to look at how you can apply their Magic Number to your own applications.
For those not familiar, the Nike+ is a sensor you put in your shoe that pairs with compatible iPod/iPhones to display stats about your run or cardio, play motivational music, and upload stats to your NikePlus.com account for analysis. The article is part of Wired’s Living by Numbers feature, which covers technology’s pretty remarkable ability to provide tools for tracking every detail of your life and using that data for constant self-improvement (sound familiar? ). Nike would love for you to buy a pair of their Nike+ compatible shoes with a sensor pocket, but you can also use Nike+ with any shoe by duct taping the sensor to the top of your shoe or buying a special pocket that attaches to your laces.
There’s a lot of really interesting stuff in there about the Apple/Nike collaboration on the iPod/Nike+ sensor and where it came from, including:
- Nike is compiling rich data about more than 1.2 million people running on their sensors. They’ve collectively tracked more than 130 million miles and burned more than 13 billion calories. This real-time data about runners has never been gathered before and is shedding all kinds of insights into the habits of that population on an international scale (e.g.: people in the US run more often than those in Europe and Africa in wintertime, the average run is 35 minutes). Air Miles and other loyalty programs work on the same principle (although their rewards are arguably much less valuable): aggregated data has value greater than the sum of its parts.
- Lesson: look for patterns that might unexpectedly emerge when you aggregate smaller pieces together. They can often inform your product design decisions much more strongly than individual data points.
- Hard data and a real time feedback loop have turned the adage “people hate exercising” upside down. This is revolutionary for anyone who builds a product in a space that people need but don’t necessarily want (versus something like chocolate, which everyone wants but doesn’t need). Technology is great at reversing long held paradigms (Wikinomics, the Long Tail, etc.), and this is no exception. This is technology mediating one human behavior (couch potativity) by leveraging another (ego flattery).
- Lesson: you can motivate people to do surprising things when the payoff is big enough, and it’s hard to get bigger than ego. Add features inspired by the world of video games to your product as a powerful carrot.
- Simple, easy to use products will always kick the ass of complex, product rich beasts. The Nike+ doesn’t include GPS or heart rate sensors, two of the most common features in other running accessories. Although most people might think those are indispensable must-haves — and Nike is thinking of adding them in later versions — the product has outsold almost everything in its market without them.
- Lesson: KISS! Keep It Simple Stupid. In the words of Antoine De Saint-Exuprey: “You know you’ve achieved perfection in design not when you have nothing more to add, but when you have nothing more to take away.“
The most interesting statistic for me is the Nike+ Magic Number: 5. To quote from the article:
Nike has discovered that there’s a magic number for a Nike+ user: five. If someone uploads only a couple of runs to the site, they might just be trying it out. But once they hit five runs, they’re massively more likely to keep running and uploading data. At five runs, they’ve gotten hooked on what their data tells them about themselves.
Here’s where we come back to user engagement. All products have a magic number — a 5 Runs equivalent — but figuring it out can be tricky. Consider this engagement model (note that all curves are approximate and not based on real math):
A quick guide:
- A: the first exposure a user has to your product. Their engagement (or awareness) is zero until that point.
- A – B: their awareness grows as they hear about your product on Twitter, see news stories about it, spot your ads in Wired, etc.
- B: user’s first actual experience with the product. They are fully engaged during this experience so make the most of it!
- B – C1: engagement degrades immediately after they stop using the product. C1 is an arbitrary point at which you can still make them a permanent user.
- C1 – D: engagement eventually returns to basically where it started at as they slowly forget all about you.
Man is that line ever depressing. But all is not lost! Consider the crazy wave coming out of C1. For most products, particularly web apps, you can turn the sinking Engagement ship around by taking action at each C point, be it a Facebook “You’ve been bitten by a vampire!” notice, a Flickr invitation to add your photos to a group pool, or a Rypple from one of your colleagues. Sustaining engagement means routinely drawing users back into your world with carefully designed touch points, timed to arrive before they drop below a C point. Nike uses the social features on NikePlus.com as that draw: emails that Nike misses you, challenges from your friends (fastest 5km run, most km over 20 days, etc.), and the ability to have your friends and family cheer you on to goals you share with them.
Let’s take a look at the curve from a Nike+ user’s perspective:
I’ve made their A-B curve a little steeper to reflect Nike’s and Apple’s combined ad budget. C1 through C5 here are basically your first five runs, interspersed with reminders from NikePlus to keep at it. Based on the article, the curve takes off sharply after that as people become heavily hooked on the stats.
What’s Your 5 Runs?
Think about your product or application. What’s your inflection point from casual to engaged users? My instinct says that it needs to be a simple number so that you can actively drive users to it; you can expend a lot of resource if you know you only need to get them to their fifth run. We’ve been spending some time figuring it out for Rypple and we have a few ideas around the number of Rypples you send, the number of replies you receive, the quality of the advice, etc. We’d love to hear from you: share your 5 Runs below. There’s value in learning from each other. Let’s discuss!