Some Work of Noble Note

May Yet Be Done

Deep Dive: ClassDojo

Leave a comment

As iconic an image of the classroom as the perfectly polished red apple on a teacher’s desk, as evocative a symbol as the moss green, lined chalkboards, is the slightly askew gold star collected by eager kindergarteners for (out)performance in the eyes of their teachers.  Why not update that for the Web 2.0 era?  ClassDojo is premised on tech-ifying classroom behavior management.  Rather than handing out gold stars or even manually tracking positive comments on a clipboard or poster, teachers can now manage all of this not just electronically but through a mobile app.  And by gamifying classroom behavior, students should become more incentivized to maximize their engagement.

For an enterprise as elementary as Gold Star Management (GSM…just kidding, we don’t actually need that acronym), the app is appropriately simple and intuitive.  It may toe the line dangerously between “feature” vs. “platform,” but there’s no denying that it does what it wants to do better than any other company out there.

The Base Case

What ClassDojo does is provide a user interface for teachers to track merits and demerits for all their students.  The teacher’s app has an avatar of all students in a given class; by simply clicking on an avatar, the teacher can assign a Plus or a Minus point based on specific behaviours.  If little Timmy is just crushing it in class, little Timmy can get a “+1” for things like “Participation” or “Teamwork.”  If little Susie is not exactly contributing, you assign her a “-1” for things like “Forgetting Homework” or being “Disruptive.”

Those merits and demerits are tracked over the course of each class and are presented in a dashboard that is intended to hover, Big Brother-like, over every single class session.  Instead of letting students maintain their own vague sense for how they’re doing in class, you give them a real-time, quantitative measure of their performance.  By displaying students as playful avatars, and giving them a score for their class presence, you do two things.  First, you remove the subjectivity of self-evaluation.  Students who think they’re doing okay but are in fact struggling (or vice versa) get the important corrective guidance to self-perception, thereby enabling them to improve.  Second, you turn class participation into a game and have students competing to get higher scores by being more engaged.

You, as a teacher, can specify merit and demerit categories.  The Company would say that this customization helps you incentivize certain behaviors.   For example, I, as a teacher, would create a separate “merit” category for Cross-Disciplinary Thinking.  Anytime a student incorporated information from another class into our discussion, I’d give him/her a “+1.”  I’d disincentivize comments that indicate students weren’t paying attention to other classmates’ comments.  This is pretty exciting.  Teachers always have their own personal pet behaviors that they want to encourage.  The problem is that, until now, encouragement could only really be a simple statement such as “I like the way you [incorporated your US History lesson]” anytime a student demonstrated that good behavior.  It’s an ephemeral and entirely forgettable reinforcement mechanism.  ClassDojo, with the bright green or red point trackers, the friendly sound that plays when awarding a “+1,” and the big text “+1 for [Creativity/Cross-Disciplinary Thinking/Teamwork/etc.]” that pops up, is an immensely powerful signifier.  Students will almost certainly forget stated objectives, no matter how verbally emphasized they may be.  Students will not soon forget formalized, concrete behavior specifications when presented in such a visually appealing, gamified context.

In addition to basic behavior tracking, ClassDojo is capable of tracking student attendance and even displaying a timer for the myriad classroom activities that are timed.  These two features are simplistic but disproportionately meaningful to the classroom environment.  ClassDojo, predictably, provides “analytics” as well.  I say “analytics” in quotes because the insights the platform can offer are quite limited.  It provides a breakdown of merits and demerits awarded on a per-student and whole class basis.  You can see whether you’ve been generally assigning more positives or negatives over time; you can see how student behavior has improved on specific merit or demerit categories; and you can track one class’ “personality” vs. another of your class’.  Here’s an example of what the “analytics” looks like.


Teachers can even prepare those participation reports on an individual student level and email them directly to parents.  I like the idea of parents being able to see their kid do better on “creativity” one week and better on “participation” the next.  It can be the start of some great dialogue at home, allowing parents to focus on improving whatever class-manifested behaviors are important to them.  All this stems from ClassDojo turning the jejune “Gold Star” into a multi-hued star family.

All of that may sound cool, but to me, it’s interesting data the same way that my Jawbone-collected sleep data is interesting.  I’m curious as to how I’m sleeping, but the data alone can’t prescribe anything around sustaining good outcomes.  One of JMI’s software cuts is “viewer vs. doer.”  In other words, does the software just show you how something is, or does it enable to you take action as well.  Obviously the latter type is better, and in the case of some LMSs, Clever, and Knewton, the data is powerful enough to make the software worthy of “doer” status.  ClassDojo is “viewer” software.

Notwithstanding the above, I should add that I discovered one great use case for this data, courtesy of a YouTube savvy teacher.  He reviews the overall class performance periodically with his students at the end of a class.  He points out how they’ve been particularly good with “participation” or “teamwork” one week, but how they’re slipping on “homework,” etc.  If a classroom is a company, this is the equivalent of providing shareholders with a 10-K or 10-Q.  This kind of quantified, objective data analysis seems really powerful to me.  In keeping with the whole gamification theme, you not only gamify individual student behavior, but you gamify overall class engagement.  It builds a great, healthy peer pressure system since students will want to improve their class’s performance and will look to the weak links to get better.

ClassDojo’s Extreme Value Case

If you ignore content management as an imperative for any real classroom domination, ClassDojo’s EVC is to serve as an almost Pavlovian (or Orwellian) behavior management mechanism.  Because every student, teacher, administrator, and parent looks to ClassDojo for standardized performance tracking, the network effect of this business is incredible.  Even if you aren’t comparing inter-teacher metrics, the fact that you can get this powerful data in one class means you demand it for every class out there.  As a result, every teacher is on this platform and every teacher uses this as the basis for class participation grades.  In fact, many teachers use it to grade on more nuanced metrics than mere “class participation.”

Class starts in the 10 seconds after every student is in his/her seat and the teacher has marked them present or absent on the ClassDojo attendance tracker.  With every student in possession of a phone or tablet, ClassDojo isn’t just a teacher-to-student messaging system, but it is capable of student-to-teacher flow as well.  There are frequent, non-disruptive ways for teachers to see how well students are paying attention through one-off multiple choice questions which teachers can craft on the fly, push out to all students and receive near instantaneous responses back.

ClassDojo is also intelligent in the recommendations it makes to improve behavior.  Whether via inter-class comparisons or time series analytics on a per-teacher-per-subject basis, it can not only track the “outcome” data like how students participated, but it can track “input” data like what types of lesson plans were used, what methods of pedagogy were employed, etc.  ClassDojo lets teachers use metadata and tagging to describe each class they ran and then spits out detailed analytics and causation to help teachers and administrators improve how classes are being taught.  Ideally, there are data scientists at every school to help each teacher build this kind of awareness.

All this success around targeted behavioral improvement makes ClassDojo so much more than just Nickelodeon-esque avatars.  ClassDojo evolves from providing value through gamification to providing value through behavioral psychology and management science.  Forget just expanding into high schools and the post-secondary world, ClassDojo is employed in human capital management across the corporate world as well.  All training sessions and even regular work meetings have active ClassDojo instances running.  Even if they don’t publicly display the point scoring, the session organizers and attendees tag input data and log outcomes for every meeting, and the Palantir-level data science that supports ClassDojo’s behavioral analytics provides constant feedback to managers and participants.


That’s a pretty meaningful leap to make for a platform that’s automating the underautomated space of gold stars.  Let’s spend a bit of time on the potential weaknesses of ClassDojo.

First, it’s possible that ClassDojo is disruptive to the classroom – and not the good kind of disruptive either but literally disruptive.  By having a class scoreboard constantly up, by giving kids a new distraction to pay attention to in class, you run the risk of losing kids attention.  Sure, the intentions are good and honest, but it’s not hard to imagine kids, especially competitive ones, losing focus on what really matters in class – learning.  It’s like when athletes start overthinking their game and, in the process, lose their effectiveness.

You could also potentially increase academic stress for students, some of whom may really hate the idea of having a constant, numerical grade publicly hovering over them.  I missed this when initially thinking through the ClassDojo value prop because of how much of a class participation animal I was in school.  (Some would say, quite disruptive in my own right.)  For people who aren’t the class participation-type, for the more reserved but equally engaged students, this is a pretty rough development.

Second, ClassDojo isn’t doing anything with school-level analytics.  I signed up as a teacher first at Patterson Park Charter and then Digital Harbor.  I could see a few other teachers at the same school who signed up but nothing around inter-class analytics or utilization tactics.  Admittedly, this functionality may be available for those who sign up as administrator; or it may be in the pipeline.  But this is a weakness or limitation on a much more fundamental level.

Why?  Utilization has to be incredibly different across teachers, so it’s possible that inter-teacher comparisons are entirely spurious.  Some teachers may award every action they observe, positive or negative, and give out hundreds of points per class.  Others may reward only the meaningful actions and only find themselves giving out a dozen or so points per class.  Others may only rarely (if ever) use the negative points to preserve its powerful reinforcement message.  In such a case, as is true of any consumer experience survey, comparing merit and demerit awarding rates across teachers, classes, and schools is an apples-to-oranges comparison.

Additionally, once you start comparing data across teachers, you incentivize them to start gaming the gamification system.  Teachers that know they’re being compared may start giving out more positives than they normally would or holding back on meaningful negatives.  There are already countless examples of teachers gaming SAT prep – in many cases, actually cheating – to make their schools look better.  When the competition is within a school or even a department, it’s hard to think that it wouldn’t intensify just based on how personal it could get.

This is just tough at its core.  Student feedback is an extremely subjective decision made on a case-by-case, per student per teacher basis.  Once you start trying to standardize it, you risk being too prescriptive and destroying the charming individuality of teaching styles that, even subconsciously, makes school more dynamic for students.  Think back on your last school experiences – weren’t there truly some days where you loved going into the high energy social studies classes vs. the low-key, Oxbridgian MCR discussion environment of your English class?  And other days where you couldn’t handle that energy and wanted nothing but the more calming class?  To think about applying analytics around behavior and classroom management could destroy that fragile ecosystem unique to every classroom.  You’d destroy the Cults of Personality that some of the coolest teachers build, and by losing that, you lose a lot of what makes the best teachers role models in the first place.

Lastly, does ClassDojo have any value outside of elementary and maybe middle school?  Look at the avatars: they’re clearly intended for younger kids.  It’s comical to think of professors deploying this in college and as you start down the grades from there, you realize you have a hard time seeing this used even in high school.  Even if you “maturify” the avatars or use profile pictures, you still hit the core issue which is the shifting expectations as students progress through their K-12 journeys.  High school was marked by the increased expected maturity.  In middle school and elementary school, students are young enough that schools and teachers devote substantial resources to just keeping kids happy to be there.  It’s why first-grade classrooms are so colorful, it’s partly why recess exists, and it’s why, at least in the US, positive feedback is such a big part of education culture.  In high school, a lot of that goes away.  You’re expected to know what the value is of an education and teachers don’t need to spoon-feed you to keep you engaged.  The focus is on conveying more content more effectively.  Does ClassDojo’s brand of gamification really have a place across the entire K-12 spectrum?


ClassDojo’s single biggest strength is its user-friendliness.  This platform makes you want to be a teacher – plain and simple.  Signup for ClassDojo is so frictionless that I pretended to run a couple of classes myself.  The power of creating my own merit and demerit categories was addictive and the novelty of handing out “+1”s and “-1”s didn’t quickly wear away during the fake lectures I gave.  (Thank you to Professor Borges for providing me with word-for-word lectures to deliver.)  In a great touch, you get an email directly from CEO / co-founder Sam Chaudhary when you sign up; this is something Buffer’s CEO, Joel Gascoigne does as well.  It’s transformative in how it energizes the user-product relationship.

Anyways, check out how great the dashboard looks.


I cannot overstate the excitement of seeing a “+1 for Cross-Disciplinary Thinking;” to me, at least, it’s truly an impactful reinforcement mechanism of the importance of learning tactic x or y.  I can only imagine the unique, valid learning practices my teachers wanted all their students to learn, but which got lost in the one-pager deluge of syllabi and course calendars that accompanied the first week of school.

Even if you lend credence to every single weakness identified above, there are a lot of benefits to ClassDojo, very few of which you need to squint to really believe.

  1. Automating the gold star awarding and management process, thereby encouraging teacher-to-student feedback.
  2. Gamifying class behavior, thereby encouraging higher quality engagement on the part of students
  3. Building a unified class mentality, thereby encouraging peer support and peer pressure to improve participation across an entire class
  4. Formalizing academic behaviors (like “teamwork” or “cross-disciplinary thought”), thereby allowing teachers to incentivize targeted learning practices, rather than leave kids with the palimpsest of all good classroom behaviors that they otherwise would get
  5. Remove subjectivity in class participation grades, if not completely, then at least partially, by having real-time updates of how participation is working and what contributes to grades one way or the other
  6. Allowing parents to get frequent, formalized updates not just on whether their kids did well/poorly in class, but the reason for that performance

The problem is whether all the above constitute a feature or a platform.  This is a problem for two reasons.  To be somewhat prosaic, “feature” apps have a limited financial outcome since larger competitors could easily provide the same functionality but with the added benefit of integration and bundling with the rest of their platform.  Less prosaic is the issue that “feature” apps have a tougher time making a real impact on their industry.  By providing just one offering – even if you do it really well – you aren’t allowing yourself to gain the cross utilization benefits that platforms have.  You also make it harder to monetize, harder to keep users, harder to scale your ambitions, and harder to gain mindshare and thought leadership status among your chosen ecosystem.  Influence is limited.

ClassDojo right now feels more like a feature than a platform.  Its best bet to move out of feature territory, I think, is through moving from mere behavior tracking to behavior management.  To use the Jawbone analogy, the band would be infinitely more useful to me if it didn’t just tell me how I slept, but facilitated my continued better sleeping.  If ClassDojo can find a way to encourage targeted academic behaviors – and by encourage, I mean more than just give kids the pretty “+1” points – it has the potential to stake out a huge chunk of the education world’s real estate.  It would find it easier to move into the corporate world and blow wide open its potential TAM.

I don’t have any brilliant, specific recommendations for exactly how to go about doing this.  I think that would call for pedagogy PhDs and significant A/B testing on different features.  But I do know that ClassDojo is set up to support this capability.  It’s too early for any comprehensive studies on ClassDojo’s effectiveness to already exist, but the Company absolutely should have plans to commission, fund, or even perform such efficacy studies itself.  It would do well to focus not just on whether ClassDojo classes have better outcomes, but in what ways those outcomes are better.  Is ClassDojo viewer or a doer software?  To return to my point above, does ClassDojo simply track behavior or can it be a tool to in fact manage and control behavior.

Parallel to all of this is the monetization question, as well.  20 million students and 2 million teachers is great, but to my knowledge (and by the Company’s admission, arguably), ClassDojo has generated no revenues to date.  Given how monadic the Company and offering have always been about their teacher focus, it’s tough to see a near-term path to monetization.  You lose teachers when you make them pay for stuff.  To state the obvious, monetization does not run counter to a mission-driven, teacher-focused culture, either.  To truly help teachers and maximize your impact on the education space, you need scale and resources.  The admirable value set of the team to date doesn’t need to be abandoned when thinking about self-sustaining product development, research, customer service, sales, etc.

So what are ClassDojo’s options?  They could go the Edmodo route and work on district-level analytics sold as a separate offering.  They could advertise, which only heightens the question of whether it would be a distraction in class.   They could strike up partnerships with educational content providers.  By this I mean not just textbook publishers but vendors of all the bric-a-brac endemic to K-6 classrooms like notebooks, crayons and markers, glue sticks, etc.  There are intelligent ways to tie those goods into prizes based on all the points awarded and I’m sure those vendors would appreciate another channel straight into classrooms.  There’s also the potential to build a teacher-generated content ecosystem, where educators can upload syllabi, materials, and tests and quizzes, selling it to other teachers or schools and providing ClassDojo with a cut of the profits.  You could tie those offerings into specific merit / demerit categories, or even build new point tracking systems and grade management tools that do command some price.  There are a lot of options, none of which are necessarily easy to start executing.  When I listed the strengths above, monetization was not one of them and this is the issue that plagues ClassDojo more than any other of-scale EdTech platform I’ve come across to date.

I’m willing to believe that ClassDojo can manage behavior down the road and that the weaknesses I’ve identified above are simply an example of luddite thinking.  Every new tech platform has negatives but the truly successful ones don’t need to address all those problems on a case-by-case basis.  They just need to add enough value on their positives, and in rare cases, do something that no one else out there can do.  With regards to managing behavior, not just in a classroom but in any supervised performance management environment, ClassDojo has that potential.

Author: AJ

I'm an education enthusiast, growth equity investor, and MBA student at Wharton.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s