The Final Touch: Trigger and Gamify

[#1]

Michael Wu, Ph.D. is Lithium's Principal Scientist of Analytics, digging into the complex dynamics of social interaction and group behavior in online communities and social networks.

Michael was voted a 2010 Influential Leader by CRM Magazine for his work on predictive social analytics and its application to Social CRM.He's a regular blogger on the Lithosphere's Building Community blog and previously wrote in the Analytic Science blog. You can follow him on Twitter at mich8elwu.

---

Welcome back. I just want to say that my first SxSW experience was a blast. Aside from all the parties, our panel was great. The audience were engaged and asked challenging questions, and I believe that is what made the panel fun and informative. If you are interested, SxSW provides a podcast of our panel on The Science of Influence. Someone from the audience also recorded a video, but the sound quality isn’t great. So if you want to watch the video, I recommend playing the podcast in the background, then play the YouTube video with volume turned off. Finally, our panel has been graphically recorded in the Ogilvy visual notes. Pretty cool huh! I am totally missing SxSW already, and I hope to attend again in the future.

By the way, if you are interested in the Science of Influence, I’m happy to inform you that I’ve been invited to speak at SugarCon 2011 on this very topic: The Physics of Influence. It will be a 3-part talk (40 min). I will cover the basics of how influence works, the ROI and some proven repeatable business strategies, and then some advance analytics involving social netowrk analysis. So I think everyone should be able to get something out of it. Hope to see you there in couple weeks.

Alright, now we’re ready for more gamification. This is the fifth article in the science behind gamification mini-series, and we’re ready to talk about the last factor of Fogg’s Behavior Model (FBM). If you missed any of the earlier articles in this mini-series, I highly recommend reviewing the following before jumping into today’s post:

1. Gamification from a Company of Pro Gamers
2. The Magic Potion of Game Dynamics
3. Psychology 101: Motivation for Gamificators
4. Simplicity Counts! Even in Gamification

Today I will talk about triggers, the final factor of the FBM. From the discussions in my earlier posts, there seems to be some confusion between trigger and motivation. Let’s see if I can make that delineation clearer today.

Why Trigger?

In my previous two articles, we talked about how game mechanics/dynamics motivate users and increase their ability (or perceived ability through increasing the task’s simplicity). This is important as we should be able to move users to the upper right of the ability-motivation axis above the activation threshold of the desired action. However, the user may still not take the action, even if he is motivated and has the ability.

There can be many reasons for the lack of action. Just to name a few examples, the user may be:

1. Unaware of his ability (e.g. didn’t know that he can take such action, unaware of the simplicity of the task etc.)
2. Hesitant (e.g. unsure if it is appropriate, unsure if it is the right time, question his motivation, etc.)
3. Distracted (engaged in another routine activity due to behavior momentum)

A good trigger is designed to solve these problems. A trigger can take many forms but its function is simple. It prompts the user for action now. The only requirements are that the user must be aware of the trigger and understand what it means.

Since the trigger tells the user to do something, it must indirectly make the user aware of the fact that he can do what he was asked to do, and it is appropriate to do it now. Furthermore, it can even serve to interrupt other routines that the user has been engaged in.

Triggers Depend on the User’s Behavioral Trajectory

Despite the simple function of triggers, there are actually many different types of triggers for people with different behavioral trajectories. This is because there are many possible behavioral trajectories to reach the activation threshold on the upper right of the motivation-ability plane.

If people already have the ability, but they are not motivated, then their behavioral trajectory may look like figure 1a. On the other hand, if they have the motivation but lack the ability, then they will probably reach the activation threshold as in figure 1b.There are also those who have the ability and motivation, but they are just waiting to be triggered to take some action.

For these three types of behavioral trajectories, Prof. B. J. Fogg has outlined three types of triggers:

1. Spark: for trajectory 1a, people who have ability, but not motivated. This type of trigger is often built-in as part of the motivation mechanism. This is why some people are confused by the difference between motivation and trigger, but triggers are not inherently tied to motivation in FBM.
2. Facilitator: for trajectory 1b, people who are motivated, but lack ability (or perceived ability). This type of trigger often simplifies the task by highlighting its simplicity. It is often used with the progression dynamics to create anticipation as the user practice and improve his ability towards the final goal.
3. Signal: for people who have both the motivation and the ability. This type of trigger should only serve as a reminder. It shouldn’t try to motivate users who are already motivated. Otherwise it can easily turn into an annoyance. It shouldn’t simplify the task either. Doing so may actually make the task seem boring and not challenging enough.

In most realistic situation, people’s behavioral trajectory will be somewhere in between these three extremes. Designing the proper triggers requires a good understanding of the users’ motivation level and ability. Although many gamification techniques focus on driving action through motivation, Prof. Fogg found that people are most receptive to facilitator and signal triggers. This is because spark triggers often aim to motivate people to take actions that they wouldn’t otherwise intend to take. This can be annoying when used in the wrong situation or too frequently.

Triggers also depend on Bartle’s Gaming Personality

The effectiveness of triggers may also depend on people’s gaming personality. As I mentioned in the introductory post to this series, early game researcher, Richard Bartle, identified four major types of gaming personalities: Achiever, Explorer, Socializer, and Killer. Different types of triggers do appeal to different personalities.

• For example, killers are extremely competitive. A trigger that challenges them can quickly launch them into action (provided that they have the motivation and ability).
• On the other extreme, socializers hate confrontation and tend to follow the crowd. So a good trigger for this type of personality would be something like “10 of your friends are doing this now, want to join?”
• Achievers are driven by status, so spark triggers that are associated with a raise of status could be very effective for this group.
• Finally, explorers are driven by discovery, the uniqueness of their contribution in a timeless and unbounded world. So a trigger that calls upon their unique skill for help without too much time pressure could be very appealing to them.

Conclusion

Today, we talked about the last factor of FBM: the Trigger. We learned that without a proper trigger, motivation and ability may not be sufficient to bring about action. Although a trigger’s function is simply to prompt a user for action, designing an effective trigger is challenging, because a trigger’s efficacy depends on many factors. We discussed two that are used often in game dynamics:

1. Behavioral trajectory: Based on the trajectory that brings a user to the activation threshold, Prof. Fogg categorized triggers into three types: (a) Spark, (b) Facilitator, and (c) Signal.
2. Gaming personality (Bartle’s gamer typology): (a) Achiever, (b) Explorer, (c) Socializer, and (d) Killer.

This post concludes our theoretical investigation of gamification. Through FBM, we learn that successful gamification is all about driving the user above the activation threshold by:

1. Motivating them
2. Increasing their ability (or perceived ability)
3. And then applying the proper trigger at the right time

This temporal convergence of motivation, ability and trigger is why gamification is able to modify, alter, and manipulate human behaviors.

Now that we have all the theory under our belt, we can start exploring the business application of gamification later. So stay tuned. I hope to see you next time. But for now, let’s have a discussion and explore the subtleties of triggers. I welcome all comments as usual.

[#2]

Hi Mike,

Thanks again for a great piece. Nice roundup.

The concept of Triggers seems to do a bit of a disservice to the FBM - at least from a purely theoretical perspective. Motivation and Ability seem elemental, but Triggers, as you note above, are not. Triggers may include a non-zero amount of Motivation and/or Ability. This compound nature seems to muddle the potentially clean and powerful formula (B=MAT).

A more parsimonious formula would result from “factoring out” the M or A from the T. In doing so we would get at the unique contribution of Triggers to the formula. In a previous comment, I suggested this component was awareness or, better, Realization. That is, what a Trigger does, uniquely, is create an immediate Realization (R) in the actor that his/her Ability is sufficient to satisfy his/her Motivation. This is what we get when we factor out the M and A from the T. I think this holds up well in light of the preceding article. If so, then the formula might be more succinctly written as: B=MAR.

Triggers (sparks, facilitators, etc) are great concepts for guiding the practical design of behavioral change solutions. However, in getting at the theoretical core I don't think Triggers go far enough because they can be broken down into more elemental parts.

Thanks,

Aaron Fanetti / @aaronfanetti

[#3]

I'd like to see specific implementations of this for say, education - what is the structure, levels, triggers, etc especially for a site that sells lessons/courses.

[#4]

Hello Aaron,

Thanks for coming back and for the detail response. I apologize for the delay response. I've been under the weather a bit lately.

You have a very good point. I think that Trigger from a implementation perspective does contain some elements of motivation and ability in it. So I agree with you that although this is very useful from a design perspective, it does lack that theoretical beauty. Perhaps this is due to the fact that motivation and ability usually takes a more continuous range of values, where trigger is binary. You either trigger or you don't.

I like you idea of factoring out the motivation and ability elements from trigger, but I'm not sure if the remaining element should be call awareness or realization. As I mentioned earlier, the function of that element is to trigger the user to take some action. And in many cases, users are very well aware and knew all along that they have the ability to take certain action, so they don't need to realize anything. However they are just not motivated.

For example, taking the trash out is something that everyone has the ability to do, but people just don't want to do it. Having some (x-factor) that make people realize/aware that he CAN (has the ability) take the trash out to satisfy his motivation just won't work, because the problem is that he is not motivated in the first place. It may even be construed as condescending.

Well, what ever we call it now, it may just be a place holder until we come up with a more suitable term for it. I'll leave that for the academics to debate and decide. This is not my theory. I simply borrow it to explain why gamification works.

But I do agree with you that from a theoretical perspective, trigger does lack that elemental essence. And thanks for the nice graphical illustration. They are wonderful. Kudos to you!

Thanks for the comment, and hope to see you again on Lithosphere.

[#5]

Hello Will,

The specific implementation is precisely the thing that I want to avoid. Because they will surely be different in every cases. Gamification is not just a single formula that you can apply over and over again. It is an iterative design process that you discover what works for your audience and optimize them through iteration. Taking a formula that work for one situation and apply it to a totally different situation is likely not going to be very effectively.

That is why my articles tend to focus on principles rather than the specific implementations. However, if you want to look at some cases, gamification.org has a growing list of gamification cases in education.

Thank you for stopping by, and thanks for the comment.

[#6]

Hi Mike,

Thank you for the response. I’m consistently amazed at the dialogue you create with your readers.

Just a quick comment on your example above. I don’t see a need to burden our X-factor with a motivational facet. Doing so just brings us right back to the original problem. The FBM already wholly accounts for the take-out-the-trash example with the Motivation (M) factor. There is no need to employ anything more. The behavior architect/game designer simply needs to increase M. The X-factor (e.g., immediate awareness) can remain unadulterated.

I think we can (and should) maintain a very hard line between all three factors – at least through an academic lens. Triggers are just too fuzzy to use gracefully at this level. Using them to guide practical solution design is no doubt a different story.

Thanks again. As always, look forward to reading more.

Kind regards,

Aaron Fanetti

[#7]

Hello Aaron,

Thank you for continuing the dialogue. I came from a strict academic training, so I enjoy these theoretical discussions. I strongly believe that is how we can all learn and advance our knowledge. So your comments are always appreciated.

So I been thinking about your model a bit. Let me know if we are thinking along the same line here.

Maybe in theory all that is needed is motivation a ability. It is conceivable that If you increase M and A sufficiently (extremely high M and A), maybe a trigger is not really necessary to bring about action. Yet, trigger appear in FBM because it is a practical design trick. It is the single component of the FBM that the designer can control fully. So in practical situation the behavior architect/game designer can use this component to bring about extra motivation or ability to push people above the activation threshold.

I think this can reconcile the fuzziness of trigger as a factor. It really should be treated as a separate design element rather than an element of the behavior model.

Anyway, just some thoughts. Thanks again for your input.

[#8]

Hi Mike,

Okay, let me go one last round. M and A alone are not sufficient. We definitely need the "x-factor". 

Consider the allegory of The Elephant and The String:

Every evening a baby elephant was tied to a stake with a string by his trainer. It tugged and pulled with all its might to break loose. It wanted so bad to be free and see the world (lots of M!). After exhausting himself time and again the baby elephant soon became frustrated and quit tugging, the string was just too strong. Time passed and baby elephant grew large (lots of A!), all the while still longing to be free. Alas, the very same string that once held a baby now held fast a massive and frustrated beast.

Our protagonist has lots of Motivation (M) and Ability (A), but something is missing. If only the poor beast knew that all he had to do was tug once more and he would attain exactly what he wants! 

Triggers may be fuzzy at a theoretical level, but they hold the essence of what is needed… awareness. Maybe it’s not so academic either. I think a behavior architect wanting to design good Triggers would probably benefit much from understanding their essential nature.

Now for an experiment, let's architect some Green Dot behavior and get Fogg himself to chime in.

Kind regards,

Aaron Fanetti

[#9]

Google Alerts triggered me here. Really.

Despite my time limitation (an ability factor), my motivation was high, so the Google Alert trigger worked: I clicked to see this page.

I have limited time right now to  scan and reply. I'll make a guess at answering without having read all the above. 

Yes, right now in my work I'm mostly concerned about help people doing real-world design. So Aaron is right here: "to guide practical solution design"

I hope I can return to this page over the weekend and read the entries more carefully.

---

Come to Stanford in April or May and let's dig into this during my class.

[#10]

Hello Aaron,

Thank you again for the interesting discussion.

I like the allegory of the Elephant and the String. 

However, it can be argued in that case there is no temporal convergence. Clearly, when the elephant was still a baby, he doesn't have the ability, even though his motivation to break free was high. When the baby elephan grew large, there can be many reasons that he doesn't break free. (a) as you said, he might not realize that he has the ability, but there can also be other reasons (b) maybe he is no longer motivated to break free anymore, even though he realizes he has the ability. Perhaps he developed a relationship with the trainer, etc. We really have no way of knowing. It can easily be argue that if the adult elephant really wants to break free, he can and he will.

As I've outline in this post, there are many reasons that motivated and capable people don't take an action. Unaware is only one of the reasons, hesitant, distracted, and there can be many other reasons.

I must clarify my self a bit. I am not saying that trigger is unnecessary. I'm just speculating that in the extreme ends, so extreme that even motivation and ability can start to trade off, then maybe trigger isn't necessary at those extremes. I definitely think that in most situation triggers are absolutely necessary. That is why I wrote this post.  :-)

I see that you are a Fogg disciple, talking about the behavior grids. It would be cool to get Fogg to chime in, but I know he's an extremely busy man.

Thank you again for the discussion. Very much enjoyed. Hope to see you again next time.

[#11]

Hello BJ,

Long time no see...

What a great honor to have your presence. Knowing how busy you are, I'm totally thrill to see a comment from you. I would definitely appreciate your comment on this subject matter, as I learned them from you the first place.

I appreciate your explanation of how you get here with your own behavior model. Nicely done and gets the point of your model across very well.

Please take your time. I am always interested to learn more from you.

Thanks again for stopping by.

[#12]

Hi Mike, I'm talking about design patterns which fit many common situations (similar to coding and enterprise design patterns), not using the same pattern across different industries. But I will look elsewhere thanks.

[#13]

Hello Will,

Welcome back and thanks for the question.

I guess I'm not very clear on what you mean by design pattern. From what you are saying here, the Fogg's Behavior Model seem to be just the kind of design pattern that you are talking about.

This design pattern basically says that you need to create a temporal convergence of 3 factors: Motivation, Ability (perceive or real), and Trigger the user when both Motivation and Ability are sufficiently high (i.e. above the activation threshold for some action). That is the design pattern, because that is what drive people to take action.

But precisely what you use to motivate, and what you use to increase the user's ability (or perception of his ability), and how you trigger him, is free for you to explore and test for your specific use case and audience.

For examples, if you have a level up scheme for status change, some times 10 levels is enough, but sometimes you need 50 levels. You don't want it to be too easy, otherwise people may feel bored. But you don't want it to be too hard either, otherwise people get frustrated. This is from the principle of Flow. The precise number of level depend on the audience and your use case. I must emphasize that gamification is not just a cookie cutter that you can just apply and expect good results. There are a lot of science behind it. That is why I decide to start this mini-series, to discuss the principles the guides the design process.

Alright, I hope I've address your question. Maybe it is still not at the level you like, but feel free to ask more specific questions here. We can always discuss.

Thanks again for the comment, and hope to see you again.

[#14]

Hi Michael,

Greatly enjoyed reading your gamification series & the various theories. 

Mathematical proofs & factoring aside, I do think that M, A, and T each have an important role to play in design and success.  Further, MAT concepts are not limited to gamification per se. 

Let us take the area of UX - website usability  & user-experience. The goal of any website is to lead visitors towards a set of conversions (could be reading content, subscribing to a feed or newsletter, purchasing a product etc.). And the sites that do this best are those that are not just elegantly designed but designed to increased the  ease of navigation & search (simplicity), but also is relevant to what the user is looking for and has elements like offers, discounts etc (to increase motivation). Finally the best conversions happen where there is a good, strong 'call to action' (trigger).  Conversely, if your conversion rate is not good (the expected behavior is not happening) then that implies one or more of the MAT elements are out of sync on your site.

In this, I would slightly deviate from Aaron’s concept of realization as the unique contribution of Triggers.  Coming from a business perspective, I view ‘triggers’ as the final psychological push that prompts someone into a decision.  In other words, more than the 'realization' about one's abilities it is a mental state that gets hot-wired into action by the combination of inputs taken in.  Again taking the website example  - having a ‘call-to-action’ is not enough.  It is the unique combination of words, color, positioning, size etc. that goes into the call-to-action that makes it a wicked trigger.

And this brings up the point that triggers cannot be generalized – and hence the importance of “personas” in web optimization. A trigger for one might not be a trigger for another and so ideally you first need to understand your segments & customer base before you start working on creating the ideal MAT combination.

Good stuff.

Regards,

Ned

[#15]

Hello Ned,

Good to see you. It's been a while since I saw you on Lithosphere. How have you been?

Glad you enjoyed reading this post, I enjoyed writing it too.

You are totally correct. Fogg's Behavior Model (FBM) is developed to guide persuasive designs. I only borrowed it to explain the efficacy of gamification. But in general FBM can be apply to any design task, beyond games and gamification.

Concerning the generalizability of trigger, even though we cannot generalize it to everyone, some trigger design principle may be generalizable to different sub populations with a certain psychological profile, such as the gaming personality that I mentioned in the post. I guess that still require you to know which gaming profile that your audience fits into. So, yes,... to design an effective trigger, you are going to know something about your audience. And the more you know, the more you can leverage those intelligence, which lead to more effective triggers.

Nice to see you again. Thank you for commenting, and hope to see you next time.