Last week, I was in Johannesburg meeting some clients, and the conversation turned toward a fundamental question – how to use technology know how for business outcomes that are real and at scale? In my view, the best way is to apply a blend of two methodologies – Data Science and Design Thinking.
Here are some compelling reasons:
- They are about “discover first” as opposed to “define first”.
- Data Science attempts to discover success criteria buried in data.
- Design Thinking attempts to discover success criteria buried in the human and systems / processes interactions.
Applying Design Thinking before Data Science would provide outputs of better quality and accuracy, which are more relevant to the personas, thus driving higher solution adoption. When kicking-off a Data Science project, begin with the end-user in mind. Use personas and stakeholder maps to understand business objectives, the barriers to adoption, and the associated risks.
Design Thinking helps data scientists to look across multiple dimensions of the problem. It not only helps them identify the potential variables that might yield better predictors of performance, but also enables them to think holistically to design the relevant solution for the end-user. The problem can be simplified by breaking it into smaller pieces: Organizational processes are complex, with lot of interdependencies, embedded business rules, situational exceptions and overrides. These factors constitute the “small data” behind the big data, and unless the problem is broken down to smaller sub-components and nuances are understood, an idealistic solution will make no business sense at all. Design Thinking helps data scientists break the larger problem into the supporting multiple use cases (identify, validate, value and prioritize) functionally, which can be quickly prototyped to validate and ideate further.
Embrace an open mind to seek new synergies and new perspectives: We can’t solve problems by using the same kind of thinking we used when we created them. Design Thinking helps data scientists to examine the underlying inefficiencies and hotspots in the processes and systems critically. The root cause may be elsewhere, the solution may be simple but with a greater risk associated with change management, or may be conflicting metrics are influencing the users to exhibit very different behavior than expected, may be some other thing, etc. Before data scientists jump to leverage their toolbox, it is extremely important to identify causal inference and the role of cause-and-effect. Design Thinking helps identify patterns, relationships and associations to establish causal inferences. One must move beyond asking questions such as what happened, to formulating hypothesis by asking questions such as what would happen based on possible interventions or different scenarios. Finally, it is important for data scientists to look out for black swans – what would happen or what outcome might have occurred if a different path had been taken.
Finally – is there a sure-fire way for Data Science / Analytics to succeed? We turn to “Six Laws of Design Thinking Inspired Data-Driven Organization”:
Law #1: It’s always about the end-user, not just the data and technology.
Data Science is all about impactful outcomes and the material differences it can bring to the business, and the most critical element in that is the end-user. The more the outcome is transforms the day in the life of the end-user, the better it is
Law #2: It’s about eliminating barriers associated with Time and Distance.
Data Science is all about becoming predictive, and hence the outcome should translate to eliminating the time and distance barriers within the enterprise and outside of it, directly
Law #3: It’s about creating new possibilities.
Data Science, beyond its current view of narrowness, and mostly confined to technology ecosystem of data, algorithms, and platforms, should pave the way to exploit and explore information to create more and better insights about customers, product, operations and markets, which can be definitely used to reimagine your business model
Law #4: It’s about continuous learning.
Data Science is all about predicting what’s likely to happen, prescribing actions, learning from the results, and integrating those learnings into your business model faster than the competition
Law #5: It’s about creating more compelling, differentiated offerings.
Once you become predictive, it is a natural progression to creating more “intelligent” applications and offerings, not only for the markets and consumers, but also for your organization
Law #6: It’s about monetization.
Data Science is about identifying, quantifying and eliminating the inhibitors to customer and market value creation, and reimagining your business processes, systems, and the business models to monetize
This is not a blog about Data Science – even the six laws do not talk about technology, cloud, ML, DL, etc. The idea is understand that we need to avoiding the temptation of putting the tech before the means, and instead, begin with Design Thinking.