Experience API [xAPI] & Why You Should Give a Rip

They say that even a blind squirrel will find a nut eventually, and I am living proof of that concept. So there I was at DevLearn 2014 locked in to speak at a couple of breakout sessions and facilitate a tabletop discussion at the crack of dawn on day one. I’m there on a Performance Support mission of “seek and learn” and purely by accident I stumbled over a pre-conference session Tuesday afternoon called the xAPI Hyperdrive. Yup…maybe not totally aligned with the squirrel metaphor, but I found a nut I was not even looking for…and it changed my entire conference agenda.

Before I go any further I think it’s important to remind my readers that the Experience API [xAPI] was formerly known as Tin Can. The question that normally follows is “So what?” and that’s a legit response because the concept of xAPI is not that intuitive until you see what it can do. I hope this post helps answer the So What question and validates that xAPI really is a big damn deal, and I’m one of a growing number of folks that thinks that it’s better to understand xAPI sooner than later.

What Is the xAPI?

In a non-technical, simplified definition…and biased by my own early sense of awe…I will describe xAPI as follows:

The xAPI is an architecture of standards by which we can track activity.

[Rim shot] Hey, thanks for reading! Are there any questions? Send cash gratuities to…seriously, xAPI is that simple. Certainly, it gets deeper when you start talking about coding and enabling API connections among disparate systems and databases, but I’m not going to risk a brain implosion by trying to swim deeper than I can. So this will be simple…I hope. The complexities should be left to the coding geeks and hackers that live outside the box and speak in code and other unintelligible dialects. I don’t have to know what they know…I only need to know one of them…and feed them…and know what I want to track…and preferably why I want to track it.

At the core, xAPI tracks activity through a protocol that was developed by ADL, the same folks who brought us SCORM. Easy now…this is not about SCORM…this is about what SCORM cannot do. Those activity statements I mentioned are based upon a simple 3-part format of:




…or to put it a different way




What I think is significant is the expanded context in which activities that involve learning can take place through a very diverse mix of experiences…outside the LMS…and we can still track them. To make that point, I’ll throw the 70:20:10 model on top.  [HINT] This really is a big damn deal…

Going back to SCORM; what do we track? LMS completion data, right? Depending on how SCORM is set up to work in your LMS, you may have visibility to other COURSE related activity, but you are LIMITED to course related activity. Period. Are those LMS-based experiences the only learning EXPERIENCES we should give a rip about? Methinks not. We have move downstream and into the post-training context; in other words, get down to the point of work. We need to include systems we work in that do not talk with the LMS. We need to link competencies with what we actually DO on the job, and we need to match performance to just how competent we are…or not…and what to do about the gaps with informed learning and performance support solutions. xAPI opens the door to a new visibility to Performance and Learning.

Again, I’ve overlayed the 70:20:10 model in the matrix below. Consider activities like these:


Experience Type





Formal Learning











This course

This module

This video

Knowledge check

Table tab #3.1


Informal Learning
Performance Support
@ Point of Work







Job aid 7.32

Task guide – NH

Media clip 2TS


Social Learning




Project Team X





Best practice

FAQ Question

Help Desk answer

Project Plan


@ the Point of Work







His sales funnel

A sale

Repair Visit

 Figure #1

Hopefully, after reviewing the matrix above you can begin to see why I’m so excited about xAPI. I added the 70:20:10 conventions in the first column just for grins…and to make the point…that xAPI is about EXPERIENCES that cross boundaries between what we track out of SCORM routines and what is additionally possible that truly defines performance at the point of work. In other words, xAPI totally supports all aspects of a Performance Paradigm as the journey to competence unfolds.

To be fair, a Performance paradigm wraps around a Training paradigm by including a holistic view of the performance and learning ecosystem. xAPI is the enabling architecture that allows us to track a more holistic picture of activities the workforce experiences in the 70:20:10 world where they LEARN and WORK. The power of this expanded view gives us is the ability to tie performance activity to competencies, and in many cases, in real time.

Following the 70:20:10 convention, we can track at an amazing level of detail, and as you see in the matrix, not just training completion data. The key, at least as I see it, is not getting too crazy with tracking things just because we can. We must answer questions around “What do we need to track in order to gauge relevant performance activities that drive business outcomes?” Obviously, there may be more questions to address, but my point is only to emphasize tracking relevant data points to match performance to results.

The kind of activity xAPI can track is a function of the different databases that are able to “see” the activity. I think it safe to say that if any system captures a record of what was done and by whom, there is enough information to build the I – DID – THIS formatted activity statement. The trick is for that database to be compatible with an API that communicates with the xAPI Learning Record Store [LRS]. Many systems already do…like major LMSs, SharePoint, LCMSs, Authoring Tools, and others. If not, the LRS vendor can often build the API as required.


Sure, why not? Actually, I think 70:20:10 gives us a framework under which we can plan what we are going to track and why specific to activity. Refer back to the matrix and look at the diversity of the VERBS; most have nothing to do with training related activities. I’ve take poetic license on my choice of verbs used in the matrix – the xAPI protocol has some structured limits, but the point remains…this ain’t about tracking SCORM-based training!

Also, note that the ACTOR is not always an individual performer/learner. It could be a work function [Help Desk] or even a group of performers [Project Team]. What we choose to track becomes a function of what needs to be tracked; the systems involved; the activities that take place; and the KPIs that are relevant measures of performance outcomes. That’s where the coder geeks come in…and they build the API connections among databases that typically do not speak to each other.

xAPI enables those conversations to take place…in the form of activity statements…and captured in a database called the Learning Record Store [LRS]. From the statements captured in the LRS we can then query, filter, slice, and dice data to feed real-time reporting venues to inform us on workforce performance and learning.

Bringing together activities from the 70:20:10 ecosystem…inclusive of “SCORM stuff”…gives us the ability to tie actual performance outcomes to competencies. What that translates to is a real-time link between performance outcome data points to competency criteria in the HRIS that can then prescribe training that resides in the LMS. We could even have visibility of who used what Job Aids [informal learning assets]that were stashed in SharePoint, or even a formal EPSS technology platform. It gets better…

Now imagine that data populating a Performance Dashboard for a District Sales Manager specific to his/her remotely deployed sale team to highlight real-time performance like…Who is doing what? When? How well are they doing it? Are there gaps? Is there training that would be helpful based upon the competencies falling short? What informal assets are being used…or not? Wow, is that cool or what?

Don’t get overwhelmed by the options I’ve tossed at you; this is only the tip of the tip of the iceberg of possibilities. The challenge will be deciding what data points need to be captured…and why. Honestly, the list is shorter when you consider what cannot be tracked than what is possible to track. Mapping the performance and learning ecosystem is going to become an art form, and it’s not going to happen in that restricted world where we hug the training paradigm to our chests like a flotation device in a water landing. We are truly staring a new, or at least evolved, set of skills like performance consulting that will be essential to thriving at the point of work.

Some Amazing Examples of xAPI Application

There were ten organizations participating in the xAPI Hyperdrive session; each had 8-minutes to share what they did using the xAPI. Prepare to be amazed. I walked out of the session muttering, “Holy applications, Batman!” I promptly changed my conference agenda to stalk anything and anyone who was going to address the xAPI.

It seems my brain is a half-duplex device. This means I can do two things…but only one at a time. Given that limitation, I can either listen; or I can take notes. Trying to do both usually means I don’t do either one well and recall suffers. Add to that a busted memory buffer and…well…here’s what I can recall that came out of the DevLearn xAPI Hyperdrive session on November 4th:

#1 – Drive Innovation in Teacher Education – Cognitive Advisors
Problem: New teachers unprepared to be effective in the classroom Application: Used the Saltbox LRS [called Wax] using the 70:20:10 learning path model of both formal and informal learning. Used xAPI to track badges earned through activities related to competence targets. Captured video uploaded via mobile for teacher mentors to review and provide feedback. Key appeared to be real time activity captured and uploaded with xAPI tracking
#2 – Emergency Needs Assessment – Catholic Relief Services [CRS] Problem: Needed to evaluate effectiveness of training for first responders real-time in a simulated disaster response Application: Learners branch through a series of simulations delivered via Cornerstone LMS. Branching awards points based upon choices made in the simulations. xAPI tracks points earned for simulations scoring. Used results to modify and fine tune training. Observations from the field will be captured and uploaded via mobile. Used Rustici’s Watershed LRS.
#3 – Connecting Adaptive & Intelligent Systems – Problem Solutions
Problem: Multiple simulations used by military not connected to enable decision making to leverage disconnected performance data Application: Interoperable Performance Assessment – needed valid data pulled real-time from disparate simulations to promote adaptive training. Training path flexes based upon real performance data aggregated from multiple simulations.
#4 – Scavenger Hunt – Saltbox
Problem: New nurses on the floor have many items needed at the bedside and must be found quickly. Application: Scavenger hunt used as a tool for tracking nurses to find what they need during onboarding activities that did not involve the LMS.
#5 – Mobile Sales Enablement – On Point Digital
Problem: WW Granger onboarding new sales reps using online portal and native apps and needed to track progress in field using iPad and IPad Application: Required tracking offline activities. The vendor built an offline player that captured activity and then synced up with LRS when a signal becomes available.LRS data enabled them to streamline learning and simplify content.
#6 – Real time KPI and Competency Evaluation – AutoDesk Problem: Preparing AD partners’ selling staff to effectively sell AD products Application: Used LRS data to measure performance, capacity and capability and the business impact of formal and informal activities. This visibility enabled quickly adapting programs based on actual performance.Data captured down to content object level…and if I recall…down to the HTML panels on a single page…very granular…and captured directly from the LCMS to map competencies to a performance dashboard format. 
#7 – Backbone for Enterprise Learning
 – Xyleme LCMS
Problem: Content being developed in many tools and formats for use in different systems of engagement. LMS, SalesForce, Social, etc. None of those systems talk to one another. Application: Needed single source content distribution to provide Smart learning content…objects with things added like competencies and using a cloud player versus a SCORM player. Used content analytics to generate a dashboard view for managers. Visibility to informal activity. Tracked below course level…enabled identifying parts of courses were good…despite the course sucks overall. Cloud xAPI eliminates need for LMS track. Can track Performance Support level usage activity by user. This one was waaay cool!
#8 – Real-Time Community Learning Environment – Instancy Problem: Community had multiple vertical disciplines and tracking member’s activity was an issue since they could be in more than one vertical. Application: LCMS was at the core. Participation was tracked by xAPI and data from LRS populated a Learner Dashboard that showed all of their learning experiences. Next phase will tack on actual performance metrics against learning experioences.
#9 – Gaming – Float Mobile Learning Problem: Teaching management concepts to mobile users Application: These guys put some amazing gaming together, and this one was called Pendulum. They combined WebGL with xAPI and used HTML5 so they had compatibility across many platforms. They were capturing actual scoring based upon attempts, successes, failures and level attainment…all of which SCORM could never do.On an related note, there was discussion later on in another session about using gaming scores at different locations to establish a little friendly competition between retail store locations at another company. xAPI brought scores and efficiency ratings and ranking together across a retail enterprise. Pretty slick!
#10 – RFED-Radio Frequency Education – Torrance Learning  Problem: Teachers needed to tie field trip learning experiences back to actual in-class curricula in order to justify taking the kids out of class and the expense of the field trip. Needed proof of learning. Application: Tracked kids activity in a Children’s Museum as they experience each exhibit and the interactive activities at each one. Each exhibit had an RFID antenna to track student interactions. Each student had RFID cards on lanyards. At each exhibit they interacted with the hands-on aspects of the exhibit and then keyed in answers to multiple choices question at each exhibit. xAPI captured exhibit activity and the answers to questions that tied back specific learning at the museum to the curriculum back at the school. 

Summary Thoughts

Looking at those ten examples, can you see the power of what’s possible with xAPI? I was blown away.

My earlier exposure to xAPI two years ago [Tin Can at the time] was primarily seeing a huge value through the visibility we could have to Performance Support usage. What a narrow view by comparison to the above case studies. Even so, that was, and still is, my mission. And that’s okay. And that’s the beauty of what xAPI is and where it may fit based upon your specific needs. Back then just knowing who was using what Job aid was enough to get me all raked up in a pile, but now…oh my!

This is a big deal…a big damn deal…and with the xAPI we are within reach of fully supporting a Performance paradigm with true real-time performance data. The Hyperdrive session illustrated just how much I do not know, so don’t feel overwhelmed by what you’ve just seen. Much like adopting an embedded performance support discipline, xAPI is also a start small and scale proposition. Look back at #4 above…a scavenger hunt? Really? Don’t try to boil the ocean even though you are within reach of a tool that is fully capable of boiling the ocean. Let’s just boil it one bucket at a time. We have to build our own competencies that support the Performance paradigm and learn what it is we don’t know about what’s possible. I find that exciting.

What I see happening is the ability for us to tie actual performance [as defined by actions taken] and map that against the HR competencies identified as relevant. Where competency gaps exist we can then have a truly prescriptive learning path plotted based upon the required performance. We will have visibility to learning performance [SCORM] as well as the informal learning assets and venues experienced by the learner/performer consumed on their path to competency. I’ve been writing about the convergence of learning with work now for a couple of years, and I’m so excited to see xAPI as the bridge between the formal and informal learning and support assets and experiences required to address convergence at the point of work. It’s so sweet to be right about something…

Gary G. Wise
Workforce Performance Advocate, Coach, Speaker
(317) 437-2555
Web: Living In Learning