Talent Code Applied

Talent Code is where Daniel Coyle describes deep learning through short repetitions and feedback loops. I have applied this approach in coaching sports and in business.

Talent Code  via Skiing with Joe

Talent Code a la Billy Kidd SkiingAn interview with Tim Ferriss is included at the end of The Art of Learning – An Inner Journey to Optimal Performance by chess and Tai Chi world champion Josh Waitzkin. Josh relates his experience skiing with ’60s Olympic legend Billy Kidd. Billy asked him:

“Josh, what do you think are the three most important turns of a ski run?”

Billy points out: “if your last three turns are precise, then what you’re internalizing on the lift ride up is precision.” We did many short traverses instead of a couple of long runs, and we reviewed each one. He was able to internalize what he did right and what he needed to improve after each fall.

Talent Code’s REPS approach can be applied in software development, and it can also grow the talent in your business / engineering organization This is referred to in The Lean Startup as the Build/Measure/Learn feedback cycle, and though the focus there is on learning, innovating and improving on customer needs, the same applies to the teams iterating through the process of finding the best way on executing delivery to that end:

Talent Code’s REPS

R – Reaching/Repeating

  • MVPs

    Developing Minimal Viable Products that you deliver to customers gets fast feedback, and you can learn and improve in small increments and the heightened importance of delivering quality to a customer can also serve as a primal cue.

  • Agile Sprints

    Iterating in short development sprints further shortens the cycle.

  • Discrete Tasks

    Breaking sprint objectives into discrete tasks that are followed by tests runs can further tighten the feedback loop.

  • Microservices

    Deploying new functionality incrementally in the form of independent micro-services also increases the ability for tighter, more focused loops, and this speeds up learning and establishing processes that are continuously being improved and optimized.
    Uber employs their Micro Deploy cycles to leverage microservices for Continuous Delivery in their application of the Talent Code.

Talent Code Cycles - Uber's Micro Deploys

Talent Code - DevOps CyclesDevOps and Microservices – Symbiotes

E – Engagement

  • Autonomy

    Don’t provide your engineers with the technical solution they implement, but rather with a clear statement of the problem. Allow them to arrive at the best solution, and this will empower them through the Multiplier approach. Providing the solution is a dis-empowering Diminisher approach.

  • Challenge and Mastery

    Engineers love to improve their craft, and if you focus first on time to delivery, you’re sending a message that time trumps quality. Instead lead them with the objective of finding the most efficient, elegant and sound solution, and you’ll feed their drive.
    Dan Pink underscores that Autonomy, Challenge and Mastery motivates people much more that monetary gain.

P – Purposefulness

  • Clear Problem Statements

    Provide a clear understanding the value to the business, and the customer will empower and motivate engineers to come up with the best solutions in striving to solve for an understood purpose.
    Dan Pink refers this as the “Purpose Motive”

S – Strong, Direct, Immediate Feedback

  • Feedback from MVPs to Customer

    Deliver MVPs and incremental improvements to your customers, and you get the fastest real feedback on how well your solutions are received by your customers. Sometimes, this can result is very strong, clear feedback that allows us to learn and course correct before we invest further down an errant path and we learn more quickly.

  • Sprint Retrospectives

    One of the most valuable aspects of doing Agile Sprints is what a team can learn from a retrospective. Here the team decides what worked well – to keep, what – didn’t – to stop, and what could be improved – to change. When I ran Yahoo!Games, I brought the release cycle down from months to releasing at changes at the end of every sprint. The learning of the loop came through adding customer feedback to the retrospective at the end of a sprint, and we then continued to tighten that loop to daily releases to production.

  • Continuous Deployments at IMVU

    When I joined IMVU, we were not only doing MVPs, but we were deploying code to production in incremental changes every 40 minutes. I helped bring that down toe 9 minute cycles, and those pushes were typically in the form of A/B experiments where we could quickly learn which were more effective at improving customer experience. We also had an Immune System which would automatically rollback changes that went out of bounds in terms of memory or disc usage, customer sessions times, etc.

  • 5-Why’s in Blameless Post Mortems

    Learning is also greatly enhanced when things go wrong and we as quickly as possible do a Post Mortem where we get to the root cause of what went wrong with the objective of learning (not blaming). This too facilitates learning through a feedback cycle, and it serves to make individuals and the team stronger. The energy right after a major issue can also serve as a primal cue to help ignite deep learning.

  • Running Tests After Each Completed Task

    A fast way to get strong, clear and immediate feedback is to break work down into discrete tasks that include tests written to stress and break the code that was just written. This can enable a feedback, learning cycle that can occur on a daily basis and it is another reason to break down work and not leaving the testing and validation to the end of a lengthy product development cycle.

  • Code Reviews on Each Submit

    Another very fast turn-around can come in the form of code reviews any time an engineer submits a change.

  • Applying The Talent Code at IMVU: Methods of Rapid and Continuous Feedback

    Talent Code Development/Deploy Cycles at The Lean Startup
    From Visually.

I should add that not everyone learning to ski would have followed me to the top of the mountain in near gale force winds that day for their second time skiing ever. Most who have heard this story tell me they don’t want to learn to ski with me. However, Joe trusted me and followed me up there, was willing to fall, get up, listen to what I had to say, and he would go again knowing full well the most likely outcome would be another fall.

Joe got up from each fall, and he didn’t think about the last or next fall even – he was thinking about getting feedback on what went wrong so he could do it better on the next stretch. Not everyone has the perseverance and courage to do that, and I give Joe a lot of credit for that. It does also help to have established a relationship of trust that I was solving for his learning and growth. That foundation of trust is vital if you hope to guide individuals into trying new things they may not be comfortable with.


The notion of many small, incremental improvements is known as Kaizen from lean manufacturing, and it also works well with the approach of many small reps with lots of opportunities for small incremental improvement also aligns well with the REPS described in The Talent Code.

John Boyd’s OODA Loop

Boyd’s “Observe, Orient, Decide, Act” also underscores the benefits of repetitive learning cycles/loops similar to those in The Talent Code

Talent Code Cycles and John Boyd's OODA Loop


See also:

Talent Code – Building Myelin 

Rapid Iteration in Software Development

Drive – What Motivates us

Multipliers – Leaders that Empower Others

Outliers and 10,000 Hours of Practice

The Talent Code provides an alternative to Gladwell’s 10,000 hour rule.

OODA (Obeserve, Orient, Decide, Act) Loop

John Boyd’s OODA Loop is another parallell to the ideas present in Coyle’s The Talent Code


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.