Subscribe to DSC Newsletter

Why do we need Learning Sprints?
Virtually every company is under pressure to transform their business in order to sustain in the future. As part of these efforts, they are hiring data scientists and engineers, data analysts, and they are making huge investments into cloud, big data technologies and AI, amongst others. Virtually all employees need to unlearn what they have assumed to be true for decades, and they have to acquire new skills. Time and money are the big constraints for any upskilling initiative.

That’s where Learning Sprints come into play as an innovative way to upskill employees.

What is a Sprint? Agile? Learning?
Before we dive into the details, let’s deconstruct the term Learning Sprint into its basic components.

What is a sprint? It’s a short, fast run with a clearly defined start and an endpoint. We want to get as quickly as possible from A to B. For example, we sprint as pedestrians when crossing a heavily trafficked road or when competing for a medal as professional runners. We never sprint around aimlessly, that’s called jogging.

The term spirit is being used metaphorically in various ways. As an example, the agile community defines a sprint as follows: “Sprint is one timeboxed iteration of a continuous development cycle. Within a Sprint, a planned amount of work has to be completed by the team and made ready for review.”

“What is agile?” someone might ask. Here’s a definition by software company  Atlanssian: “Agile is an iterative approach to project management and software development that helps teams deliver value to their customers faster and with fewer headaches. Requirements, plans, and results are evaluated continuously so teams have a natural mechanism for responding to change quickly.”

Last but not least: “What is learning?”. Here’s what Wikipedia says: “Learning is the process of acquiring new, or modifying existing, knowledge, behaviors, skills, values, or preferences.”

Now that we have outlined a rudimentary definition of “sprint”, “agile” and “learning”, you might conclude: “Aren’t we stating the obvious here?” The answer is: “Yes, we are!” However, we cannot assume that everyone is on the exact same page. Corporate history is full of grandiose failures which all start with the lack of common understanding of the most basic terminology and its meaning.

Definition of a Learning Sprint
To sum things up, a Learning Sprint is a collaborative, short, fast-paced learning endeavor as part of a greater, agile-oriented team effort. This being said, if you speed read a book alone at home, that’s not a Learning Sprint. Don’t confuse “learning fast” with a Learning Sprint. Also, a Learning Sprint is something you do with others, not alone.

Now that we’ve established a common ground, let’s grind into the details.

How can you facilitate a Learning Sprint at your company?
Here’s what you need:

Context
First of all, you have to define the scope of your learning efforts in the context of a concrete project at your company. This might be a greater, high-level initiative or a specific task, which requires new skills. The more specific you can be about your context, the better. A Learning Sprint is not an isolated island.

Objective
What is it that your team wants to learn? Be realistic about the goal of a Learning Sprint, and also be specific. Here’s a weak objective for a Learning Sprint: “We want to understand machine learning.” Here’s a stronger one: “We want to get our hands dirty with TensorFlow 2.0 and with Keras. Our objective is to gain a foundational understanding of these frameworks, how they democratize the usage of machine learning and how we can possibly leverage on those technologies to accomplish [enter your business objective here].”

Don’t be too ambitious with your Learning Sprint. Realistically, if you miss your learning objective by 30%, bear in mind that a 70% accomplishment is a good result. The deeper you dive into a subject, the more you become aware of the breadth and depth of the topic of your choice. It’s nearly impossible to hit a Learning Sprint objective by 100%.

Preparation
Learning Sprints take a decent amount of preparation. First, you (the Learning Sprint team) have to define the learning objective, which should be aligned with a specific business goal. Second, you need to reverse engineer a viable learning path for a Learning Sprint to accomplish that goal. What sources of knowledge do you aim to leverage on?

Whatever you want to learn, there is a high chance that some sort of content is already available out there. It might be a course on a MOOC (Massive Open Online Course) or … simply a video series on YouTube. Leverage on it.

Team
Who is going to be part of your Learning Sprint? In agile, team size matters. Too big teams don’t get too far, because people stand in each other’s way. Too small teams, by contrast, might implode before they even hit the ground and run. A team size of 5-7 people is just perfect.

What is the composition of your Learning Sprint team? A diverse team needs a different approach than a homogenous one. As an example, you can host a Learning Sprint on TensorFlow 2.0 with a pure-play data science team or with a diverse team. In each case, the learning objective and the learning path is a different one.

Autonomy
Don’t confuse a Learning Sprint with instructor-led training. A Learning Sprint is a collaborative, autonomous learning effort. Do you need an expert on a Learning Sprint team? Not necessarily. First, experts rarely possess the skill to translate their knowledge into a language digestible for non-experts. Second, as stated above, a Learning Sprint is a collaborative effort. With an expert on board, the remaining team members would tend to lean back and passively consume someone else’s knowledge.

A Learning Sprint can be characterized as learning by doing, not as learning by listening. You might, however, sit down with an expert after a Learning Sprint and get his feedback on what you have learned.

Time
A Learning Sprint might take anywhere between one and five days. Less than a day makes no sense, neither does more than a full workweek. When in doubt about the duration, make a shorter Learning Sprint. Learning is hard. Normally, it’s harder than regular work, because of the cognitive overload. Our brain needs a lot of energy to process, map and integrate new learnings with the knowledge we already possess.

Also, a Learning Sprint is not an around-the-clock Hackathon. Realistically, in a corporate environment, we can’t stay focused on one specific learning subject for more than six or eight hours straight. Participants often state that an hour on a Learning Sprint feels almost like two hours of regular work.

Focus
Humans are bad at multitasking. It decreases our productivity, and it can also damage our brain. Being on a Learning sprint means: No emails, no phone calls, no interruptions, no meetings. Not even digressions into related subjects.

A Learning Sprint on machine learning is a Learning Sprint on machine learning. It’s not a Learning Sprint on the ethical implications of technology in general. If your team has a tendency to digress, consider forking a separate Learning Sprint.

Structure
Make sure you structure your Learning Sprint for maximum effectiveness and everyone’s preferred learning style. Don’t over-structure it though. You don’t need to plan every minute of your Learning Sprint.

Things might change during the Learning Sprint slightly, and your team should have the freedom to make moderate adjustments on the flight. After all, we are practicing agile, right?

Communication
The goal of a Learning Sprint is to learn something new and to generate shareable knowledge. Make sure you make time to communicate your insights after the Learning Sprint with a broader audience within your company. These can be executives, peers or colleagues in different business units.

Consider following up on your Learning Sprint with an internal webinar, a Q&A session, an internal blog post or a combination of the above. Since your Learning Sprint is aligned with a specific business goal, being public about your learning efforts is not advised.

Repetition
Think long term. How many Learning Sprints do you need to acquire substantial knowledge of a subject over the course of a year, two or maybe even three years? Also, don’t rest idly between Learning Sprints.

A team aiming to organize multiple Learning Sprints should determine what learning efforts are needed between each sprint. Split up responsibilities. For example, a team of five can read five different books in the weeks or months between two Learning Sprints.

Freedom
A Learning Sprint is not about force-feeding employees with knowledge. If someone shows up in a Learning Sprint because “my boss sent me” - send him back to work. A Learning Sprint requires everyone’s maximum intrinsic motivation. Nobody should feed on someone else’s motivation. That’s called energy vampirism, and it’s not a good thing.

Respect someone’s decision not to participate. Also, raise the question: “What other topics might interest people apart from that of our planned Learning Sprint?”

Creativity
Learning Sprints thrive on creativity. Take risks. If you want to improve the quality of decisions being made on your team, you might consider a Learning Sprint on chess. If communication is an issue your team wrestles with, you could make storytelling the subject of your Learning Sprint. Some of the world’s greatest authors and screenwriters share their knowledge online, for example on Masterclass. A team of data scientists who master the skill of storytelling can be of great value to any company.

Don’t limit yourself to the next best subject on your list. Be creative. Take a Learning Sprint as an opportunity to explore unknown territories instead of just manicuring a well-gardened area of expertise.

I run a Data & Technology Literacy Consultancy. Please leave a comment, or reach out to me on LinkedIn.

Views: 493

Tags: Agile, Analytics, Data, Learning, Machine, Science, Sprint, TensorFlow, Upskilling

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Videos

  • Add Videos
  • View All

© 2020   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service