Game Thinking: Innovate smarter & drive deep engagement with design techniques from hit games

by Amy Jo Kim

Speed bump #2: Relying on extrinsic motivators If your core learning loop design relies too much on extrinsic rewards and motivators, you may be able to drive short-term engagement, but your house of cards will be likely to crumble.

Speed bump #3: Forgetting to close the loop I’ve worked with many startups and game studios who can create a compelling activity, but have a hard time with driving reengagement.

Systems Explain the World in Simplified Form

In social gaming, we have Bartle’s player types: achiever, explorer, socializer, and killer. Richard Bartle, an early MUD developer, noticed that certain social patterns emerged in a variety of game settings. He described these patterns as player types, and laid out an analytic framework for thinking about online social behavior.

Things get tricky when you take a model developed for one purpose and context, and transplant it elsewhere.

I found that Bartle’s model works well for multiplayer social games like MMOs, MOBAs, and arena shooters, but less well for casual games, educational games, and health games.

Kim’s Social Action Matrix Inspired by Bartle, I took my experience with designing social games and identified four actions, or verbs, that emerge in online environments: compete, collaborate, explore, and express.

Compete This is the urge to test your skills and see how you stack up. Competitors find ranking systems and zero-sum mechanics appealing, because those structures mirror their internal dialogue and POV.

Collaborate The urge to work with others toward a shared goal. Collaborators love the feeling of “winning together.”

Explore This is the urge to gain knowledge, explore boundaries, find loopholes, and know the rules that govern a space.

Express This is the urge toward self-expression and a hunger to personalize their experience, make their mark, and express their uniqueness.

Here are some common actions mapped against the four quadrants. Which of these actions can your customers do in your product experience? Map those out, and you’ll see your social engagement style.

Prototype Your Core Activity, Not Your Marketing Message. Make sure that your team knows the difference between product discovery and product experience. A landing page will test your marketing message, but it’s a waste of time if what you really need is to validate your core product experience and value proposition.

Embrace a “Good Enough” Visual Aesthetic You Can Update Quickly. To maximize learning, build mock-ups that are good enough to test and learn from. Leverage your early adopters; they don’t need fancy visuals to offer useful feedback.

Use Prototyping Tools to Quickly Bring Ideas to Life. Find the quickest, simplest way to visualize and test your ideas.

An effective prototype can take many forms. You can productively test: a competitor’s product that you want to understand better. a scenario walk-through, illustrated with sketches, mock-ups, or wireframes. clickable mock-ups (built with a prototyping tool) with which someone interacts. a simple working prototype, website, or crude early version of a game. a hardware prototype that your subjects can touch and interact with.

If you’re early in the design process, you may have sketches, wireframes, or mock-ups of your ideas. Now, not everyone has the imagination to look at low-fidelity visuals and envision what they might become.

If you want early customer feedback on a product idea, UI flow, or progression system, testing low-fidelity visuals can be highly instructive and can save you time and headaches down the road.

Cheap, fast iteration is powerful because you fall in love with your own idea if you spend too much time articulating it.

A great way to visualize and test your core learning loop is with a few clickable mock-ups of key pages and flows. Many product creators use prototyping tools like Keynote, Invision, or Balsamiq to quickly mock up and test different versions of their product.

Clickable mock-ups are great for getting feedback on your UX flow, screen design, and details like navigation, naming, and layout.

A manual-first test of your value proposition is called a “Wizard of Oz” test (if the humans are hidden) or a “Concierge” test (if the humans are visible). This can be a highly effective way to connect with your early customers and find out what they want, before you’ve invested heavily in engineering. Airbnb and Aardvark are two famous service-based apps that started as manual-first MVPs.

Select the prototype format you will use at this stage of development. Sketches or static mockups Interactive mockups (computer, phone, tablet) Paper prototype (mostly for gaming) Hardware prototype Simple, stripped-down working game, website, or system Concierge/Wizard-of-Oz service

Speed bump #2: UX and graphics perfectionism Watch out for the lure of visual quality and polish. There’s a time and place for that in product development. During early prototyping, polishing your graphics and UX will slow you down dramatically.

If you’re building an app, product, or game that’s aimed at consumers, you will probably want to pay your subjects. A good default rate is $50/hour, but I’ve paid anywhere from $25 to $100 for a one-hour interview, depending on the age group, income level, and expertise level of the participants.

By default, most people think of user testing as a solo activity and will schedule one person at a time for sessions. If your product is likely to be used solo—without social interaction—this is a good approach, because you’re reproducing your core use case.

Pair testing can be especially valuable for surfacing people’s thought processes, because your subjects will naturally talk to each other as they interact.

Part 1: Warm-Up In the first part, you’ll get to know you customer better and set them at ease. Which questions should you use to open the testing session?

Part 2: Test Next, you’ll show them your prototype or product (in whatever form), let them interact with it, and notice their reactions. This might be:

a competitor’s product (to get insights into your market). an early hardware prototype (duct tape and components). a playable paper or physical mock-up. a working prototype of a website, app, or game. clickable mock-ups, sketches and wireframes.

Give them tasks and ask them to “think aloud” The ideal way to run a playtest is to give people a challenge and watch them figure it out. Try to create tasks for your customer that test your assumptions.

Question 1: Habits and Triggers When would you see yourself wanting to use this product? What time of day? What location? What would you be doing beforehand, and right afterwards?

Question 2: Likes and Dislikes What did you like most about this experience? And what didn’t you like? There are no right or wrong answers. We want your blunt, honest opinion.

Question 3: What I’d Improve What would improve the experience for you? If you could wave a magic wand, what would you love to see that would really impact your experience and make your life better?

Use Interviewer/Notetaker Pairs You’ll get more out of these sessions if you run them with a colleague. Have one person conduct the interview and guide the subject through the testing session while the other takes notes and identifies emergent patterns.

What were the key findings? What did these people like and dislike? Start by sifting through the data and noting the top 3–5 patterns from your sessions, paying particular attention to habits, ideas, and unmet needs that surfaced. When

Which hypotheses did you validate? Which ones need revising/revisiting?

What were the most surprising things you learned? If you’re surprised by your findings, pay attention! That’s learning in action. Take note of ideas and insights from your research that were unexpected and relevant.

Speed bump #1: Putting words in their mouths During the interview, your job is to understand your test subject’s mental model and opinions. It’s tempting to jump in and put words in their mouth. RESIST.

As you’re prioritizing your results, focus on your early adopters, the people who are excited to get their hands on what you’re offering and will put up with pain and friction to get it.

Pattern #1: The Fashion Browser

Pattern #2: The Co-Creator One play pattern we surfaced is the co-creator, who loves to dress, shop, and primp with a buddy, usually a close friend or relative.

Just like in real life, you can raid your friend’s closet and borrow her things if you’re friends in the game. This creates an incentive to build your friend network, because it expands your wardrobe and chances of putting together a winning outfit.

Pattern #3: The Armchair Stylist Another common pattern we surfaced was the armchair stylist, who’s highly opinionated about fashion and loves to tell other people how to dress.

Speed bump #1: Scope creep Watch out for the natural tendency to expand your scope and broaden your goals—especially if you got confirmation on your product ideas from early customers.

Speed bump #2: Lost in the details Your project brief is a summary of actionable, relevant info—NOT an exhaustive list of every detail you’ve heard and learned. Too much documentation is a sign of weak pattern recognition and makes it harder to glean the important takeaways.

Empathize with a Few Hot-Core Early Adopters, AKA Superfans. Paul Buchheit says it best: build something a few people love, even if most people don’t get it right away.

Create a Pleasurable, Repeatable Activity that Fits Their Existing Habits. Once you’ve empathized with superfans, you use them to design your MVP or alpha around existing their habits and needs. You build a simple, compelling learning loop around your core systems and then leverage your superfans to give your feedback, validate your ideas, and bring those systems to life.