Anyone who has ever participated in running a Design Sprint can attest to how fast-paced and result-oriented that experience is. When Jake Knapp at Google Ventures published Sprint – How to solve big problems and Test New Ideas in Just Five Days, it brought Design Sprints into the business horizon. “The big idea with the Design Sprint is to build and test a prototype in just five days. You'll take a small team, clear the schedule for a week, and rapidly progress from problem to tested solution using a proven step-by-step checklist. It's like fast-forwarding into the future so you can see how customers react before you invest all the time and expense of building a real product”. At Digitalist, we've conducted and participated in over 100 Design Sprints. As a result, we've had many well-structured debates around the methodology.
Questions like “Should a Sprint be run over the course of a week or longer?” or “can the five days be cut down to three?” have been discussed and tried out. At Digitalist, we take a co-creation and research approach to developing solutions. Because we've learned to run Design Sprints very well, we help businesses bring their ideas to life, ultimately helping lay out the plans that help businesses in their hope to disrupt the market.
We're often asked about how to get started with Artificial Intelligence and Machine Learning. The common approach is to spend 6 months cleaning essential data, analyze and hypothesize its impact on the product and then at that point, begin strategizing. While in some scenarios this may be required it is not the fastest or the most efficient way to start. Yes, Deep Learning and Neural Networks are mind-blowingly powerful and require good quality data. However, very few organizations have the discipline or opportunity to begin cleaning up their data strategies when the end goal is still unknown.
..This (AI/ML) Sprint process works… It empowers you with the knowledge of how and where to invest in AI in your company, and where it can be used to take advantage of your opportunities or solve your most pressing problems.
Which leaves us to ask: Is it possible to solve the problem by approaching AI challenges from a Design Sprint perspective? Could the goal of getting a clickable prototype in front of end-users be valuable enough that it justifies taking a few shortcuts along the way, such as using smaller data sample sizes?
Our answer: Yes!
It's important to understand that in every full-blown Design Sprint we've experienced, there are significant learnings that people discover together:
Our experience running AI/ML Sprints has been positive from both ours and our clients' perspectives. The sprints we’ve conducted with our clients, often in the form of workshops, have resulted in some eye-opening moments. We've discerned over the big questions like "can this really be done?", then pull out all the stops to create prototypes and revel in the adrenaline rush when it’s presented to the users.
It's imperative to be careful with setting the right Design Sprint goals. It may not be feasible to create a prototype within 24-36 hours, and it definitely isn’t something you can go live with your actual customers. However, when it comes to tackling new ideas, a 'From Ideas to Life' concept, using this Sprint process works. It seeds an awareness and excitement of AI in your employees and unearths valuable insights into your organization's future capabilities. It empowers you with the knowledge of how and where to invest in AI in your company, and where it can be used to take advantage of your opportunities or solve your most pressing problems. And all this with a cost effective, relatively small investment in a 5 day AI/ML sprint.
The answer was best expressed by one of our participants:
"Typically, this amount of decision making and shared organizational learning would have taken us 6-8 months, now it took two weeks in calendar time!"
Curious to know whether an AI/ML sprint could work for your current project? Let us help you figure that out.